Page 1
University of Michigan Health System
At a Glance
An academic health system within a
major public research university
UMHS Hospitals and Health Centers
o 817 beds, 44,000 admissions
o 1.6 million out-patient visits, 100
clinics
o 18,000 employees
UMHS Medical School
o 1,600 faculty physicians
o 1,000 resident physicians
o 700 medical students
Fighting Cancer with Linear Accelerators and Accelerated Processes
On a bookshelf in the office of Dr. Theodore Lawrence, head of radiation oncology at the
University of Michigan Health System (UMHS), is a May 5, 1958 cover of LIFE magazine about
new cancer treatment methods that shows a patient about to receive a beam of radiation. Draped
in a white sheet and lying on a hospital treatment table, the patient is dwarfed by an x-ray
machine the size of a medieval siege cannon.
Since then radiation technology has changed dramatically. Today‟s machines, called linear
accelerators, are much smaller and much more accurate. They deliver a very precise dosage to a
small area of a patient‟s body, killing cancer
cells but harming fewer healthy cells. But the
processes surrounding treatment have not
kept pace. In fact, they often are
overwhelmed by the increased complexities
in technology and escalating demand for the
latest treatment.
“Radiation therapy today is almost
unrecognizable compared to 1958,” said Dr.
Lawrence. “The ability to target and deliver
radiation has increased amazingly. Processes,
however, are almost the same.”
Most of the steps in radiation therapy revolve
around three major processes. In order of
occurrence they are: consultation with a radiation oncologist, simulation and planning of
treatment, treatment. With the exception of a couple steps in planning, which now use
computerized axial tomography (CAT) scans to create 3-D images of organs, the basic process
today would be familiar to the Life patient.
How long it took her to go from first step to the start of treatment is lost, but in mid-2005 as part
of an effort to expand a lean transformation, a study by the UMHS radiation oncology
department discovered that only 43% of new patients with bone or brain metastases were
receiving consultation, simulation, and first treatment within a day. For most, first treatment
required three visits to the hospital over five days and could take as many as 10 days. Typically,
patients came in for a consultation, went home, came in for simulation, went home, and then
came in for their first treatment.
lean.org
Page 2
lean.org © Copyright 2008 Lean Enterprise Institute 2
Beginning treatment as soon as possible is important for patients with brain or bone metastases, a
condition in which cancer cells are spreading. Radiation therapy relieves pain and preserves
bone and brain function. “This is a group of patients who really need help,” Dr. Lawrence
explained. “They‟re having changes in mental status from the brain metastasis, they‟re having
severe headaches, they‟re having pain in the bone. Everybody in the department knows these are
people who have to be treated as quickly as possible.”
The department already had found a way to begin treatment very quickly for another group of
patients. People with spinal cord tumors began radiation therapy within hours after consultation.
“They need immediate treatment or they could suffer paralysis,” explained Dr. Lawrence. The
accelerated procedure delivered treatment quickly, but it was inefficient because it relied on
working around or expediting the normal steps in consultation through first treatment. He
thought a fundamentally better process could be designed using lean principles. “We said, „Let‟s
try to use lean thinking tools and philosophy to design a system for patients with brain and bone
metastases where we don‟t have workarounds or expediting.‟ ”
Ultimately, a series of cross-functional lean improvement teams designed and implemented a
much-improved process so that the percentage of patients receiving consultation, simulation, and
first treatment on the same day rose from 43% to nearly 94%, a level the department has
sustained. “We now do same-day treatment essentially 100% of the time for patients who want
it. That is about 94% to 96% of our total patients,” said Kathy Lash, department director of
operations.
And the teams did it without hurling expensive new technology at problems. In fact, some of the
biggest advancements came from relatively simple tools such as value-stream mapping and
standardized work that improved communication between steps. These advancements helped the
team achieve a second major goal set by department leadership: improve the morale and work
life of department staff who were working longer hours due to an increasing patient workload
and increasing complexity in treatment technology.
Background The introduction of lean thinking at UMHS followed a path tread years earlier by many
manufacturing companies: it began with a quality improvement effort.
“I‟d say the quality improvement movement in healthcare began heating up in the 1980s,” said
Dr. John Billi, associate dean for clinical affairs, University of Michigan Medical School, and
associate vice president for medical affairs, University of Michigan. Government reports
criticized healthcare for quality while industry complained about escalating costs. As hospital
managements in the U.S. discovered the work of quality guru W. Edwards Deming, they began
adopting total quality improvement efforts.
But the efforts came in “fits and starts,” said Dr. Billi. U.S. hospitals initiated scattered projects
to improve an emergency department or operating room. After a few months of progress,
systems often regressed to the old ways of working. Such efforts represented “improvement as a
side activity not improvement as a basic business strategy,” said Dr. Billi. Furthermore, these
early efforts often involved nurses and staff but not doctors, which hurt sustainability.
Page 3
lean.org © Copyright 2008 Lean Enterprise Institute 3
At UMHS, when a key leader of the quality effort left in the late 1990s, the effort lost steam.
“We still continue to do quality improvement projects in every department but the concept of an
overarching improvement program that was pervasive throughout the organization was lost,”
said Dr. Billi.
Then in 2003 an executive from Virginia Mason Hospital and Medical Center in Seattle, WA,
who was a UMHS alumnus, spoke at the school about Virginia Mason‟s application of the
Toyota Production System as an overarching business system to improve hospital processes. “I
was personally very impressed, transformed you might say,” Billi recalled. That evening he
went to a bookstore to buy Lean Thinking. For the next six months, he read about lean
management and talked to “everyone in the country” applying lean principles to healthcare.
Dr. Billi‟s research led him to a Michigan neighbor and a proposal. GM, burdened with
burgeoning healthcare costs, offered to show UMHS managers and staff its lean efforts to help
them learn the concepts. The car maker also loaned the hospital a group of experienced lean
coaches to train UMHS staff and assist with six projects. Rather than create a separate kaizen
promotion office, the hospital organized its lean effort under the existing quality improvement
structure, called the Michigan Quality System, to integrate both activities to create an
overarching continuous improvement philosophy.
The six lean projects gave UMHS a cadre of eight in-house coaches from a variety of
backgrounds, including engineering, nursing, and quality. In addition to their regular duties, they
helped hospital departments with lean projects. Besides valuable training, the early projects with
GM also gave the hospital a desire to go solo. “The facilitation from GM gave us the advantage
of using experienced facilitators,” said Dr. Billi. “But the disadvantage is that you can become
dependent on the outside facilitators.”
Scoping a Pilot Project Like Dr. Billi, Dr. Lawrence had been studying the application of lean principles to healthcare,
particularly how it could help radiation oncology. “Radiation oncology really turns out to be a
great lean test bed,” Dr. Lawrence said. We really have definable steps that are well laid out: we
see the patient, plan the treatment, treat the patient.”
In July 2005, Dr. Lawrence and about 20 department managers and staff participated in a high-
level kaizen workshop that targeted the entire department for improvement, which would have
covered all types of patients and cancers. Afterwards, leadership decided to narrow the focus to
a pilot project aimed at treating within 24 hours patients with brain or bone metastases, who
represented 15% of the roughly 1,550 new patients treated annually by the radiation oncology
department‟s site at UMHS.
“We scaled down the initial project scope,” explained Lash. “We thought we would fix
everything but realized the project was too broad. We decided to tackle some of the smaller
pieces and learn more about lean tools as we went.”
Tightening the focus of the project had several advantages:
Page 4
lean.org © Copyright 2008 Lean Enterprise Institute 4
Starting first treatment within 24 hours for this group of patients would relieve their pain,
while preserving bone and organ function.
A same-day start of treatment would decrease the number of visits patients had to make,
an important factor for many patients who lived outside southeastern Michigan where
UMHS is located.
The medical information necessary to initiate same-day therapy would almost always be
available because nearly 100% of patients with brain or bone metastases were referrals
from UMHS physicians or clinics.
Procedures for the planning and delivery of treatment to this group of patients were well-
understood and accepted by department faculty and staff.
The pilot and subsequent work by lean teams also would help department leadership address the
need to improve capacity and staff morale. Lash, who also is a registered radiation therapist
(RTT), said today‟s advanced radiation technologies have improved treatment but added time
and complexity to the department‟s workload that stressed capacity and, ultimately, staff.
Ten years ago, she noted, machines delivered radiation from one or two sides of a patient. The
typical treatment session lasted about 15 minutes. Today‟s radiation machines are controlled by
multiple computers using imagery and lasers to deliver high energy x-rays from five or six
angles synchronized, in some cases, to a patient‟s breathing. The therapy is far more precise and
spares more healthy tissue. But therapy sessions, which now last from 15 to 90 minutes, are far
more complex and time consuming to setup.
By mid-2005, treatment often went beyond the time allotted for a patient, which meant that a
radiation machine and its usual complement of two therapists were delayed in treating the next
patient, an effect that snowballed through the daily schedule. This led to long delays for patients
and long days for staff members, who often worked past the normal 6 p.m. quitting time until 9
or 10 p.m. to treat everyone scheduled. It also meant that physicists, who make sure the
department‟s five linear accelerators are available and capable of delivering a beam of radiation
to the right tissue at the right dosage, had to wait until late at night or early in the morning to
perform regular maintenance. The long and unpredictable hours caused low morale and
turnover, especially among radiation therapists.
Radiation therapist Patrick Clark noted that many staff members had an hour drive home after
leaving the hospital late at night, compared to working at medical clinics where work usually
finished at 3:30 p.m. “After a couple of years that looks a lot more attractive,” he said.
Thus, the pilot project would be much more than an effort to improve efficiency or capacity. It
was, as Lash noted, an attempt to learn “how do we change the culture to introduce standard
work so our professional processes are stable.” In addition, based on what they had learned
about lean management, leadership believed that easing the overburden on staff and equipment
was essential to maintaining high-quality and safe care.
Page 5
lean.org © Copyright 2008 Lean Enterprise Institute 5
Diagnosing the 7 Wastes Here are examples of Toyota‟s
classic seven wastes from
radiation oncology:
1. Correction
Fixing an incomplete
medical chart;
2. Overproduction
Setting up a treatment
room for a patient who is
late or has canceled;
3. Motion
Searching for charts;
Walking to retrieve
equipment from a
different treatment room;
4. Material Movement
Getting equipment from a
different treatment room;
5. Waiting
Patients waiting for a
delayed treatment;
Waiting for a doctor,
radiation machine, or
information;
Unevenness in the arrival
of patients and treatment
plans;
6. Processing
Making unnecessary
movements in the
treatment room;
Redundant or unnecessary
mental or physical work;
Therapists scheduling
patients for future
appointments;
7. Inventory
Obsolete forms;
Information or material
waiting in queue.
Understanding the Current State
The department established a 16-member lean team representing all front-line service providers,
including clerical staff, attending and resident physicians, nurses, radiation therapists, simulation
therapists, administrators, support staff, and physicists.
Guided by lean management‟s emphasis on understanding
the current situation based on facts, rather than opinions or
assumptions, the team developed over a 90-day period: a
current-state value stream map of the existing process for
treating patients with bone and brain metastases; a leaner,
much-improved treatment process represented by a future-
state map; and a detailed work plan that assigned members
specific tasks with timelines for making the proposed new
process illustrated by the value-stream mapping a reality.
The team calculated key metrics for the current-state map:
Process time: the actual time it takes to complete an
activity;
Total lead time: the total elapsed time to complete an
activity;
First time quality: the probability that a patient will
go through all individual process steps without
encountering a quality-related problem;
Process cycle efficiency: process time divided by
total lead time to measure what percentage of time is
spent in value-added and nonvalue-added activities.
Nonvalue added activity, or waste, is any activity that
consumes resources but creates no value from the
perspective of the customer.
The resulting current-state map revealed lots of waste in a
therapy process requiring 27 individual steps for
consultation, simulation, and first treatment. Process time
(or value-added time) for the whole process averaged little
more than half a work day (290 minutes or 4.8 hrs.), based
on a 10-hour day. However, the total lead time for patients
to complete the entire process was often seven days in real
time (10,000 minutes), resulting in a value-added proportion
(process cycle efficiency) as low as 3% (290 divided by
10,000).
The team calculated that the entire process contained 7,825
minutes of waiting time between steps and that only 0.2% of
the bone or brain metastases cases went through the entire
process without the need for some sort of rework. Only 43%
Page 6
lean.org © Copyright 2008 Lean Enterprise Institute 6
of the patients in the six months before the pilot project completed the entire series of steps in a
day.
As it drew a leaner future state based on data and direct observations, the team and leadership
zeroed in on problems to attack in order to create a process that would begin treating patients on
the same day. Since the current-state map showed that radiation therapy had less than five hours
of actual process time “we thought we should be able to deliver treatment in a day,” said Dr.
Lawrence. “Certainly if we saw a patient before noon, we could deliver treatment by 6 p.m. if
we got rid of the waste and focused on the process time.”
The team developed a leaner future-state map based on these key observations: A lack of
standardized work meant information flow into and within the department was chaotic, leading
to poor communication, low first-time quality, and delays triggered by searching for paperwork,
equipment or the need to fix missing or inaccurate information, such as unsigned forms or wrong
details in charts. This condition necessitated the practice, which is common to healthcare in
general, of inspecting work as it was passed from one step to the next.
The leaner future state process performed consultation, simulation, and treatment within a day by
making all needed information available at the start of the process, implementing standardizing
work, and applying clear process guidelines. These actions cut overall process steps to 16,
reduced process time to 225 minutes, shrank total lead time to a day, and created the potential to
bring first-time quality up to 100%.
The team decided to begin implementation by developing standardized work for consultation, the
first step. Subsequent lean teams developed standardized work for the two other major steps --
simulation and first treatment -- and attacked other problems surfaced by the value-stream
analysis.
First Step: Consultation
The therapy process begins when the radiation oncology department receives requests for
treatment, usually from other UMHS departments. During consultation, patients are examined
by a radiation oncologist, who often is assisted by a resident doctor and a nurse, to confirm the
diagnosis and discuss a treatment plan with the patient.
The initial lean team, along with the affected clerical and professional staff, developed a standard
form, called a consultation scheduling form, to be used by referring hospital departments and
radiation oncology doctors who received referral calls from other doctors. The form, which
could be emailed or faxed to doctors outside UMHS, was explained to clerical staff members. It
required such key information as:
Patient name, address, social security number, gender, phone number, insurance
coverage, and primary care physician‟s name;
Referring physician name, unique identity number, clinic name, office address, phone
number, and contact name;
Diagnosis (determines whom the patient will see in the radiation oncology department) ;
Medical information, including:
o Office notes
Page 7
lean.org © Copyright 2008 Lean Enterprise Institute 7
o Pathology reports
o Radiology reports
o Surgery reports
o Laboratory results
o Treatment summary, if treated before
Besides using the new consultation form, the department implemented new standardized work
that required radiation oncology schedulers to quickly check an online calendar to identify what
doctors were available to see bone and brain metastases cases within 24 hours. Schedulers called
patients to tell them what times were available.
When a consultation time was set for that day, the new standardized work directed schedulers to
notify the nursing desk assistant, who notified the chief radiation therapist and billing department
that a bone or brain metastases case had been added to the day‟s schedule. If the referral came
from a UMHS department, the nursing assistant also called radiology for the patient‟s images
and retrieved medical documents from the in-house database for the medical chart being
prepared for consultation. The chart was prepared in a standardized format based on input from
doctors.
During the course of establishing standardized work, the team realized that clerks needed
training in medical terminology so they could recognize when referring doctors were describing
brain or bone metastases cases that had to be scheduled immediately. “We had to really educate
them in medical terminology,” said Lash. “We had not allowed them to be the thinkers they
really were. We had put them into robot mode.”
Simulation
After consultation, a patient has his or her treatment planned and simulated, a step that needed
overtime virtually every day to complete its schedule. So, after the first lean team completed its
work on consultation, a second team tackled developing standardized work for this critical step.
The main issues for this phase of the project were:
Reduce the long waits patients often had in the waiting room before treatment
because previous appointments ran late.
Keep treatments to the allotted times so delays don‟t mount, extending operating
hours.
Improve employee satisfaction by working normal operating hours.
The team received valuable help in observing actual work and gathering additional data from
several University of Michigan engineering students familiar with lean principles. Team
members, including the engineering students, observed treatments at each of the five radiation
machines, focusing primarily on the duties radiation therapists performed. They also observed
front desk clerks scheduling consultations and staff and doctors working on the simulation step.
This familiarized everyone with the overall process of delivering therapy so they could break it
into logical tasks, which were recorded on the current-state map along with important data, such
as cycle times and first-time quality.
Page 8
lean.org © Copyright 2008 Lean Enterprise Institute 8
The next step took nearly three weeks. Beginning in early October 2006, team members
observed 20 patients during their first three treatment sessions. The team selected these first
sessions because they set the course of subsequent therapy sessions, which could continue for
several weeks. With clipboards and observation sheets in hand, team members followed patients
from the time they arrived, noting when each task began, when it ended, and if any rework,
equipment problems, or unusual incidents occurred. They observed staff performing such tasks
as: educating patients about the treatment procedure, helping patients onto treatment tables,
checking medical charts, activating the radiation beam, recording dosages given, helping patients
out of the treatment room, and scheduling the next appointments.
The team also distributed surveys to radiation therapists, dosimetrists (who calculate the
radiation dosage prior to treatment), clerks, and physicists to discover what employees thought
caused delays. After analyzing the surveys, team members conducted follow-up interviews with
a sample group of employees to clarify the top causes.
During simulation, a patient lies very still on an examining table while a radiation therapist uses
a special x-ray machine to define the exact place on the body where the beam will be aimed.
Simulation may use CAT scans or other imaging studies to help the oncologist and therapist plan
how to direct the radiation. The simulation may result in some changes to the treatment plan to
spare the greatest possible amount of healthy tissue from receiving radiation.
The body area receiving radiation is marked with a “tattoo,” tiny dots made by a temporary or
permanent marker showing where to aim the beam. Depending on the type of treatment, the
radiation therapist may make body molds or other devices that keep the patient from moving
during treatment. These are made from foam, plastic mesh, or plaster. In some cases, the
therapist will make shields that cannot be penetrated by radiation to protect organs and tissues
near the treatment field.
Information from simulation, along with the radiation oncologist‟s guidelines about radiation
dosage, is downloaded to computers in a planning room, located near the treatment rooms, where
dosimetrists use formulas to calculate radiation dosages.
The initial study had noted that a lot of information needed by simulation was communicated
verbally by phone or in notes resulting in miscommunication, rework, lots of checks to catch the
miscommunication, and when missing information caused delays “people pointing their fingers
at each other,” said Lash.
Typical delays included: patients arriving at simulation with unsigned consent forms, patients
arriving late because they were delayed by a prior chemotherapy treatment, and therapists calling
doctors to check information. Radiation therapists had a basic idea of how doctors wanted
patients setup for treatment based on the type of cancer. They‟d page doctors for verbal
instructions to find out if there were exceptions, for example, if a patient couldn‟t lie down. If
the doctor was with another patient, simulation could be delayed.
The team first devised a few easy-to-use standardized forms to obtain the information that
simulation needed and that subsequent steps needed from simulation. The documents,
Page 9
lean.org © Copyright 2008 Lean Enterprise Institute 9
essentially checklists, were designed to make sure necessary information was collected and
conveyed simply and clearly from step to step -- from referral call to consultation, from
consultation to simulation, from simulation to first treatment, and from first treatment to
scheduling subsequent treatments. It‟s an approach Dr. Lawrence called “focusing on the white
space.
“A lot of people want to burrow into how they do their step,” he explained. “We do some of
that, but most of our gains have not been inside the steps, they‟ve been in getting rid of the white
space, the handoffs. It‟s how do you get from step A to step B, not so much what happens inside
the space, although we have had some nice gains there too.”
What follows are descriptions and then samples of the main forms designed and refined over
many months by lean teams working on consultation, simulation, and first treatment:
Patient Activity Documents collect key information during consultation that is needed by
subsequent steps. Doctors and patients simply check boxes corresponding to standard
procedures and write in any exceptions. The front of the form is for doctors and
therapists to enter information about treatment (see page 11). Patients complete a basic
medical history on the back (page 12).
Physician Simulation Orders (page 13), also are filled out by doctors during consultation.
They give therapists at the next step -- simulation -- essential information, such as
whether or not the treatment area has to be immobilized. If a doctor wants the standard
setup, he or she simply checks a box on the form. If there is an exception to the standard
setup, for example, if the patient can‟t lie down, the doctor writes in the new instructions.
“It goes through a logical order of what we want to do with the patient, we sign the form,
and it‟s done. Today‟s work is done today,” said Dr. Lawrence. He added that the
simulation orders also are good teaching tools for doctors to use with assisting residents.
After starting by creating Simulation Orders for brain and bone metastases cases, the
department developed forms covering treatment for 90+ cancer types, such as breasts and
esophagus. For example, the Simulation Orders and Planning Directive below are for
breast treatments.
Treatment Planning Directives (page 14) completed by doctors after simulation, this form
goes to the planning room where dosimetrists use it with formulas to calculate dosages.
The Patient Treatment Scheduling form (page 15) is filled out by radiation therapists and
goes with the patient after first treatment to a centralized scheduling office where a
scheduler establishes the rest of the treatment schedule.
Page 10
lean.org © Copyright 2008 Lean Enterprise Institute 10
Spreading the Transformation Under the umbrella of the “Michigan Quality System,” the University of Michigan Health System uses
lean thinking concepts to create a consistent approach to quality and process improvement. Besides the
application of lean principles in radiation oncology, other lean projects include:
The Emergency Department used value-stream mapping to cut 10 minutes off the patient discharge
process, which allowed two extra patients to be admitted daily. And ED second and third year
residents now must do problem-solving projects using the 5Ss the 5 whys.
Managers and senior managers are using A3 reports to foster dialogue and build consensus for
solving problems, making proposals, and reporting status in budgeting, strategic planning, IT
planning, operating room management, ambulatory care improvement, inpatient discharge, and
other areas.
In the Catheterization Lab a lean team standardized the activation process to reduce the “door to
balloon” time from when a patient arrives to when a balloon or stent is inserted to clear a blocked
artery. The lab averaged 72 minutes in 2007, compared to the American Heart Association‟s
national goal of 90 minutes.
The Pathology Lab, using work flow analysis and material handlers (“water spiders”) following set
routes at regular periods to collect blood samples, reduced the time to deliver samples from 32
minutes to nine. Other improvements cut analyses turnaround times by 38%, travel distances by
33%, and waiting times by 78%.
At the new Cardiovascular Center one of 17 lean projects underway standardized surgical
preparation processes in the Electrophysiology Lab so surgical patients are properly prepared for the
procedure 100% of the time.
Page 11
lean.org © Copyright 2008 Lean Enterprise Institute 11
Page 12
lean.org © Copyright 2008 Lean Enterprise Institute 12
Page 13
lean.org © Copyright 2008 Lean Enterprise Institute 13
Page 14
lean.org © Copyright 2008 Lean Enterprise Institute 14
Page 15
lean.org © Copyright 2008 Lean Enterprise Institute 15
Page 16
lean.org © Copyright 2008 Lean Enterprise Institute 16
“You want to make the forms very simple and look for the exceptions,” explained Lash, who
said the forms enhance safety. “I tell people all the time when they say, „Cathy why do we have
to use the standardized forms or do any more lean projects?‟ The bottom line is for safety.”
First Treatment The last form -- Patient Treatment Scheduling -- represented a breakthrough in how radiation
therapists worked. The major tasks for therapists, who usually worked in teams of two but
sometimes three, were basically to help a patient onto the treatment table and into any
immobilization devises. Then, from a control console outside the treatment room‟s heavy steel
doors, they used computers and imagery to align the radiation beam with the patient‟s tattooed
treatment area, and activated the radiation beam. After treatment, a therapist helped the patient
off the table, then the other therapist would schedule the patient‟s next appointments. Therapists
also were responsible for taking calls from patients who wanted to reschedule appointments. In
the earlier survey, therapists had identified scheduling as their leading cause of delays, which
was supported by observations.
Scott Hadley, a clinical physicist and member of the team, said, “As we collected data and
watched what went on, everyone noticed that after first treatment a therapist sat at a computer
with the patient spending 15 or 20 minutes trying to schedule the next several weeks of
treatment, working around vacations, work, or chemotherapy appointments while the other
therapist needed help in the treatment room getting the next patient started.”
“We decided to standardize scheduling and take it out of the control room, which keeps the
therapist focused on treating patients and puts the scheduling in the hands of people who are
really good at scheduling,” said Lash, who developed guidelines for a central scheduling clerk to
follow. The new Patient Treatment Scheduling form (above) gave the scheduler key information
for developing a realistic schedule, including the type of machine needed for treatment, time
needed for treatment, and whether or not the scheduler must coordinate radiation therapy with
the chemotherapy department.
“Taking scheduling out of the control room solved so many problems,” said Lash. “We stopped
the chaos that it caused and patients liked it much better. It‟s a little more comforting for them to
come into the scheduling room away from the treatment area,” she said.
Box Score for Treating Patients with Bone and Brain Metastases
July 2005 Dec. 2006
Consultation process steps 27 16
% patients scheduled and
treated on the same day 43% 94%
Avg. Overtime/mo. 22 hrs. 0 hrs.
% of patients arriving on-
time for simulation 22% 73%
Maize and Blue
The initial lean team also had called attention to data it collected during work observations
showing that the times for therapists to perform the same tasks ranged from 1 to 7 minutes. The
Page 17
lean.org © Copyright 2008 Lean Enterprise Institute 17
discrepancies, due to inconsistencies in procedures, made it difficult to accurately predict how
much time was needed for treatments.
“We saw that there was no standard work for who was doing what,” Lash said. Department
leadership, along with therapists, developed two sets of standardized work with primary and
secondary tasks. One set is for therapists working in a treatment room helping patients onto the
treatment table and into an immobilization device. The other is for a therapist assigned to the
control console preparing the program, aligning the machine, calling up patient records, and
reviewing patient charts for additional instructions. (Standardized work also was developed for
when there are three therapists assigned to a machine. See sample blank forms below.)
Each day, therapists are assigned to a specific team and wear a colored shirt (maize or blue for
University of Michigan colors) to identify whether they are at the console or treatment room.
Having standardized work ensured that each treatment followed the same procedure and that a
specific person was accountable for completing each step accurately. The standardized work
advanced continuous improvement. By having therapists administer treatment identically, areas
of waste became easier to identify, countermeasures became easier to develop, and the
effectiveness of the improvements became easier to track.
Std Work - 2 RTTs Std Work - 3 RTTs
Task RTT1 RTT2 Task RTT1 RTT2 RTT3
x o x o *
x o * x o
o x o * x
o x x * o
o x o x *
x o
x = primary o = secondary * = tertiary
Billing Guidelines By January 2007, oncology department leadership was extending the lean transformation beyond
the main treatment processes to supporting tasks, such as video teleconferencing, purchasing,
and billing, where mistakes in coding bills were causing a lot of rework.
Page 18
lean.org © Copyright 2008 Lean Enterprise Institute 18
When staff members delivered any procedures to patients, such as x-rays, they entered billing
codes on patients‟ medical charts. Codes for complicated procedures such as simulations and
actual radiation treatments varied with levels of complexity and radiation dosages.
Chief radiation therapists and supervisors audited every patient‟s medical chart at the end of
treatment to ensure all charges were captured and that billing codes were correct. On virtually
every chart, they found errors such as missing codes or wrong codes that needed correction.
Lash said few inaccurate charts ever went to insurance companies for payment, but the mistakes,
auditing, rework, and correcting the wrong bills that did slip out, all resulted in a major time sink
for her and her supervisors, who reviewed at least 1,600 charts annually for an average of 20
minutes per chart.
After training in lean concepts and studying how the billing process worked, a lean team
identified root causes for the problems preventing 100% first time quality, including:
Lack of training for new employees.
Lack of training on the computer system used to enter codes.
Interruptions due to heavy patient volume.
Information about coding mistakes was not being fed back to staff. “They didn‟t know
they were doing anything wrong because we never told them. We just fixed the charts,”
said Lash.
Over the course of nine months, the team developed billing guidelines then ran training sessions
for everyone entering codes. The training is given to new employees and repeated annually for
everyone entering codes. The billing software was changed to make it more user-friendly and to
catch potential errors by prompting anyone entering a charge not usually associated with a step.
Finally, supervisors fed information about errors that they caught back to staff members.
With supervisors now spending just four minutes auditing charts for correct codes, the
department is much closer to its goal of eliminating auditing, Lash said.
Box score for Billing Process Improvements
May 2006 Feb. 2008
Avg. time to audit a chart 20 min. 4 min
1st time quality (% of time that
all charges are 100% correct,
the first time)
0.11% 95.3%
A Different Kind of Kaizen The billing lean team, like the earlier teams in radiation oncology, continues to meet to improve
the process they “own” and check that past improvements are sustained. Billing is one of six
cross-functional lean teams currently operating in the department. While their approaches to
training, data collection, mapping, and use of lean tools would be recognized by Lean Thinkers
in any service or industrial company, teams have had to modify their kaizen workshop design to
accommodate the realities of a hospital environment. Unable to shut down the department or
even an area for several days to implement improvements, as is the norm during most kaizen
Page 19
lean.org © Copyright 2008 Lean Enterprise Institute 19
workshops, radiation oncology teams meet every other week for two to two-and-a-half hours to
further the kaizen process.
“In between meetings we have homework to do, such as data collection, that we report on at the
next meeting,” explained Lash, who serves on each team and often gets involved in data
collection and implementation in order to tie the entire departmental effort together. Intervals
between meetings are used to implement improvements approved at team meetings.
Reactions to Change
As improvements took root and processes improved, initial skepticism about lean thinking
withered.
“In the beginning I wasn‟t always the easiest one to work with because I really didn‟t understand
the reasons,” said radiation therapist Melanie Hamilton, who admits she wasn‟t fond of
collecting data. “In fact I was resistant to some of the methods because I didn‟t understand how
they would help until we picked a process, worked on it, and I got to see it all the way through.
Then I recognized that this would be really good for all of us.”
And staff also quickly recognized the benefit of lean process improvements for patients. “If you
can take somebody who is in pain and instead of five days to get started we can treat them on the
first day, that‟s wonderful,” said radiation therapist Phil Zegarowski.
As continuous improvement becomes the normal operating procedure, Lash is beginning to see
her role shift from traditional management based on firefighting to lean management based on
problem solving and mentoring. “I love problem solving,” she said, “and I‟ve always loved it,
but I used to problem solve by putting out fires, but they would pop up somewhere else. Now
when someone comes to me with a problem, they know they should have already asked the five
whys. I help guide and mentor people so I‟ve become more of a coach rather than a firefighter.”
In the early days of the department‟s lean effort she saw her workload increase as she was
transitioning between firefighting and coaching. “Now I‟m starting to feel the relief,” she said.
Getting Docs on Board
Dr. Lawrence said the lean transformation also is changing his approach to managing. “I used to
think that a chairman‟s job was to give answers. It isn‟t. These people know their work better
than I do. What I‟ve learned is that I‟m good at asking questions but not good at giving
answers.”
He also learned that big projects aren‟t always needed for big gains. “We made some initial
errors in trying to scope projects that were too large. So now what we do is develop value-
stream maps, identify small problems, and pick them off. Find a bottleneck, fix it -- okay what‟s
the next bottleneck? Form little teams, do little things, but remember that those little things
incrementally pile up. As you pile up a bunch of successes, people get the spirit.” Including
doctors.
Page 20
lean.org © Copyright 2008 Lean Enterprise Institute 20
Doctors initially resisted the lean effort, arguing that “we‟re a great department why do we have
to change?” or “the current process works for me,” Dr. Lawrence recalled.
“I kept telling them, „I‟m not telling you how to practice. We‟re just standardizing how you‟re
going to communicate.‟ “But in an academic research organization, standard work is a complex
undertaking. We hire the best researchers and now we‟re telling them to be standardized. You
have to show them that if they do this work in a standard way, it will free up time for research.”
He also signaled to staff in leadership positions to serve on lean teams. “It‟s now part of being a
leader in this department.”
A big breakthrough in attitudes as well as process improvement came with the successes of the
brain and bone metastases pilot and subsequent improvements in the treatment planning process.
Doctors and staff realized the lean effort wasn‟t an attempt to change care giving but the
processes supporting it. Now the department is poised for another breakthrough, according to
Dr. Lawrence.
As a result of studying the work in the simulation step from a lean thinking perspective,
improvement team members realized that patients requiring very sophisticated planning and
patients requiring simple planning could be separated. New simulation software packages and
some additional training will allow radiation therapists to plan the simple cases, giving highly
trained dosimetrists more time to concentrate on the complex cases and saving time for patients
with simple planning needs.
“This is a real job challenge for the therapists,” said Dr. Lawrence, “but they are excited about it
because it will help them grow professionally. And it will off-load a task from the dosimetrists.”
When the lean transformation began two-and-a-half years ago, radiation oncology was treating
about 120 patients daily until 9 p.m. or 10 p.m. Now it treats the same or more by 6 p.m.
Besides the resulting benefits of greater capacity, speedier treatment for patients, and less
stressful days for staff -- all of which were expected -- Dr. Lawrence was surprised by an
unexpected benefit.
“What I didn‟t anticipate were the benefits of a calmer atmosphere and how that would improve
staff morale,” he said. “Chaos just drains your energy. When you feel like you have to redo
things and develop workarounds, it takes the peaks out of our day. It drains your energy when
you feel like you are fighting the system. Being able to treat this number of patients and not
have to redo work -- you can just feel the morale improve. Things are calm.”
And a relapse to the old way of working doesn‟t appear likely. “I see a lean process in
everything now,” said Dr. Lawrence. “We‟ll do this forever.”
For More Information: University of Michigan Health System -- UMHS includes three hospitals, approximately 40
health centers, 120 outpatient clinics, the U-M Medical School and its Faculty Group Practice,
the U-M School of Nursing and the Michigan Health Corp. UMHS was honored by the
American Hospital Association for highest quality care and patient safety. UMHS was the lone
Page 21
lean.org © Copyright 2008 Lean Enterprise Institute 21
finalist for the AHA-McKesson Quest for Quality Prize. U-M Hospitals and Health Centers was
named one of "America's Best Hospitals" for the 13th year in a row and received recognition for
excellence in 15 areas of specialized care in the 2008 U.S. News and World Report.
The Michigan Quality Health System -- MQS is part of the continuing effort to improve quality,
safety, efficiency, and appropriateness across the U-M Health System‟s three missions of patient
care, education, and research.
See also: Presentation by James Womack, founder and chairman, Lean Enterprise Institute:
“Lean Thinking: From Factory to Health Care & Tools to Management”
Lean Enterprise Institute -- LEI is a nonprofit education, publishing, conference, and research
organization with a mission to advance lean thinking around the world. LEI runs monthly
regional workshops on basic and more advanced lean tools. You can read complete descriptions of
workshop content with the latest dates and locations at LEI‟s education page. LEI workbooks and
training materials -- all designed to de-mystify what a sensei does -- show you what steps to take on
Monday morning to implement lean concepts. Visit the LEI product catalog to see the resources
available for supporting lean transformations. Learn about creating lean enterprises at the next Lean
Transformation Summit.
Glossary (Adapted from the Lean Lexicon, available in printed and electronic versions)
Lean Thinking
A five-step thought process proposed by Womack and Jones in 1996 and the title of their book
that described the thought process as a way to guide managers through a lean transformation.
The five principles are:
1. Specify value from the standpoint of the end customer by product family.
2. Identify all the steps in the value stream for each product family, eliminating whenever
possible those steps that do not create value.
3. Make the value-creating steps occur in tight sequence so the product will flow smoothly
toward the customer.
4. As flow is introduced, let customers pull value from the next upstream activity.
5. As value is specified, value streams are identified, wasted steps are removed, and flow and
pull are introduced, repeat this process again and continue it until a state of perfection
is reached in which perfect value is created with no waste.
(Adapted from Womack and Jones 1996, p. 10.)
In 2007, Womack and Jones simplified the five steps to these—Purpose, Process, People:
Purpose: The primary purpose of any organization and first step in any lean thought process is to
correctly specify the value that the customer seeks in order to cost-effectively solve the
customer‟s problems so the organization can prosper.
Process: Once purpose is clarified, focus on the process (value stream) used to achieve this
objective. This is generally the combined result of three processes: product and process
development, fulfillment from order to delivery, and support of the product and the customer
Page 22
lean.org © Copyright 2008 Lean Enterprise Institute 22
through the product‟s useful life. These primary processes are made possible by many
secondary, support processes inside the organization and upstream.
The ideal process is one in which every step (action) is:
• Valuable: Meaning that the customer is willing to pay for the step because it creates value and
would object if the step was deleted.
• Capable: Producing a good result every time.
• Available: Being able to operate whenever needed.
• Adequate: Having the capacity to keep production in continuous flow.
• Flexible: Permitting a range of products within a product family to move through a process
without batching and delays.
In addition, in the ideal process the steps are linked by:
• Flow: So the good or service proceeds immediately from one step to the next without stopping.
• Pull: So the next downstream step obtains just what it needs from the next upstream step when
continuous flow is not possible.
• Leveling: From some pacemaker point to smooth the operation of the process while still
addressing the needs of the customer.
People: After identifying the primary and support processes needed to create value for the
customer, make someone responsible for each value stream. This value-stream manager must
engage and align the efforts of everyone touching each value stream to move it steadily toward
the customer while elevating performance from its current state to an ever-better future state.
Doing this requires:
• A master plan for the enterprise, often called strategy deployment.
• Frequent improvement cycles for each process, often performed with A3 analysis embodying
value-stream maps.
• Standard work with standard management for every step in each process.
Overburden
Refers to one of the “three Ms,” -- muri -- meaning overburdening. The three Japanese terms --
muda, mura, muri -- are often used together in the Toyota Production System to collectively
describe wasteful practices to be eliminated.
Muda: Any activity that consumes resources without creating value for the customer.
Within this general category it is useful to distinguish between type one muda, consisting
of activities that cannot be eliminated immediately, and type two muda, consisting of
activities that can be eliminated quickly through kaizen.
Mura: Unevenness in an operation; for example, a gyrating schedule not caused by end-
consumer demand but rather by the production system, or an uneven work pace in an
operation causing operators to hurry and then wait. Unevenness often can be eliminated
by managers through level scheduling and careful attention to the pace of work.
Muri: Overburdening equipment or operators by requiring them to run at a higher or
harder pace with more force and effort for a longer period of time than equipment designs
and appropriate workforce management allow.
Page 23
lean.org © Copyright 2008 Lean Enterprise Institute 23
Seven Wastes
Categorization of the seven major wastes typically found in mass production. Formulated by
Taiichi Ohno, the Toyota executive widely credited as the chief architect of the Toyota
Production System.
1. Overproduction: Producing ahead of what‟s actually needed by the next process or
customer. The worst form of waste because it contributes to the other six.
2. Waiting: Operators standing idle as machines cycle, equipment fails, needed parts fail to
arrive, etc.
3. Conveyance: Moving parts and products unnecessarily, such as from a processing step to
a warehouse to a subsequent processing step when the second step instead could be
located immediately adjacent to the first step.
4. Processing: Performing unnecessary or incorrect processing, typically from poor tool or
product design.
5. Inventory: Having more than the minimum stocks necessary for a precisely controlled
pull system.
6. Motion: Operators making movements that are straining or unnecessary, such as looking
for parts, tools, documents, etc.
7. Correction: Inspection, rework, and scrap.
Standardized Work
Establishing precise procedures for the work in a process. Standardized work, once established
and displayed, is the object of continuous improvement through kaizen. The benefits of
standardized work include documentation of the current process for all shifts, reductions in
variability, easier training of new operators, reductions in injuries and strain, and a baseline for
improvement activities.
In a production process especially, it is based on three elements:
1. Takt time, which is the rate at which products must be made in a process to meet
customer demand.
2. The precise work sequence in which an operator performs tasks within takt time.
3. The standard inventory, including units in machines, required to keep the process
operating smoothly.
Value Stream
All of the actions, both value-creating and nonvalue-creating, required to bring a product from
concept to launch and from order to delivery. These include actions to process information from
the customer and actions to transform the product on its way to the customer.
Value-Stream Mapping
A simple diagram of every step involved in the material and information flows needed to bring a
product or service from order to delivery. Value-stream maps can be drawn for different points
in time as a way to raise consciousness about opportunities for improvement. Value-stream
maps show the entire team all the steps involved in the work in order to improve the whole
process, instead of just optimizing individual steps. A current-state map follows a product or
service‟s path from order to delivery to determine the current conditions. It reveals all the
actions and processes required to deliver a service or product to a customer. A future-state map
deploys the opportunities for improvement identified in the current-state map to achieve a higher
Page 24
lean.org © Copyright 2008 Lean Enterprise Institute 24
level of performance. It outlines a leaner process delivering value to the customer faster, with
fewer defects, using fewer resources. A future-state map is constructed by determining if any of
the steps in the process can be done with less waste, or can be eliminated altogether. In some
cases, it may be appropriate to draw an ideal-state map showing the opportunities for
improvement by employing all known lean methods including right-sized tools and value-stream
compression.