|
Process Improvements boost
quality and physician satisfaction at Christian Hospital
Customer-InspIred® Six
Sigma methodology
reduces report delivery time by 50%
by Doris Peckron, Director,
Health Information Management,
and Marty Herbst, Manager, Transcription
Christian Hospital faces
increased competition from other St. Louis area hospitals and, in a
proactive move to cut losses and improve goodwill with admitting
physicians, instituted a process improvement initiative that focuses on
the primary customer of each key process. One such process undergoing
transformation is the delivery of transcribed reports – a process that
touches every patient going through the hospital. This article describes
how turnaround times were slashed in half and what further improvements
are planned.
Background
Christian Hospital is a
non-profit organization, 493-bed acute-care facility located on 28 acres
in unincorporated north St. Louis County, Missouri. A founding member of
BJC HealthCare, Christian Hospital has more than 600 physicians on staff
and a diverse workforce of more than 2,500 health-care professionals who
are dedicated to providing the best care using the latest technology and
medical advances.
In an area of St. Louis
where a lot of older residents live, 55 percent of Christian Hospital’s
patients are on Medicare. This large, aging population comes in with
multiple issues, which require more specialist consultations before
surgery, and places a heavier burden on internal services within the
hospital. In late 2004, Christian Hospital retained Shaw Resources to help
launch a process improvement program to improve service efficiency,
bolster patient and physician satisfaction, and boost quality throughout
the organization.
Problem
The Chief Financial Officer
receives a monthly report from the Transcription Department regarding
history and physical, consult, and operative transcription report
turnaround times and error rates. Even though the department was
consistently meeting its service guarantee of 24-hour turn around times
(TAT), the CFO was concerned about physician satisfaction with report
availability. Additionally, the CFO was looking at current patient
admission/discharge cycle times and thought that an improvement in
transcription TAT would improve physician satisfaction by providing them
with patient information in a shorter timeframe.
Process Improvement Approach
In June, 2005, the
Transcription Department was asked to participate in the Process
Improvement initiative to address report turnaround time. A
cross-functional team was assembled with representatives from each of the
affected departments: Marty Herbst, Manager, Transcription, Doris Peckron,
Director HIM, Judy Moore, Hospital-Based Transcriptionist, Diane Williams,
Home-Based Transcriptionist, Jackie Hoffman, Clerical Support, and Jerry
Lohman, IS. Under the guidance of James Shaw, an outside consultant, they
began gathering the data required to put together a process improvement
plan.
With the patented
Customer-Inspired® methodology formulated by Shaw Resources for a
framework, the team followed a continuous process improvement structure
similar to the Define Measure Analyze Improve Control (DMAIC) structure of
Six Sigma, with one caveat: everything
had to be addressed from the point of view of the customer of the process.
Define the Process
Using the commercially-obtained
Process Advisor® software, the team worked through an eight-step patented
process definition methodology, as follows:
-
Beginning with a process purpose
statement, which explains why the process exists within the
organization, the team brainstormed and established their Charter
Statement: To provide timely, accurate reports to the proper location.
-
Next, a process owner had to be
identified, someone who coordinates all process activities and is
ultimately accountable for process performance. Marty Herbst, Manager
Transcription, took on this role.
-
Working to refine the process
definition, the team composed a table of outputs – what the process
yields – along with identifying who the primary customer of each
output is and what measures best indicated the delivery of the output
to the customer. In this case, the output table is as follows:
Figure 1: shows the
Outputs table produced by the Process Advisor software.
-
Once the outputs and
primary customers were determined, the team then established the end
of the process as experienced by the customer – when the report is
placed in the chart. Customers of a process can be internal to the
organization, like the physicians who dictate the reports, or they can
be external to the organization, such as a nursing home or other
facilities. All the customers of the reports, in the case of Christian
Hospital, are caregivers: whoever needs the reported information to
make a decision about the care needed for that patient.
-
Working in reverse,
the team then identified the inputs – the products and services or
information needed to produce the outputs of the process – and who
the suppliers of the inputs are. Here, the suppliers are Treating
Physicians who furnish complete and understandable dictation.
Figure 2: Shows the
Inputs table for the Process Profile® Graphic.
-
Once the inputs and suppliers
have been pinpointed, the beginning of the process, from the point of
view of the primary customer, is determined – in this case, it is
when the caregiver completes dictation of a report.
-
To clarify what happens in a
process, a meaningful name is assigned. The Customer-Inspired
methodology calls for a verb-plus-object format when naming a process;
our team came up with the name " Provide
Transcription Services Sub-Process."
This process name identifies all reports and tables associated with
this process improvement project.
The final step in Defining the
Process brings everything together in a high-level diagram that
summarizes the process. Using the patented software, the team produced
the following Process Profile® graphic:
Figure 3: The Process
Profile Graphic documents the definition of the Provide Transcription
Services Sub-Process.
A significant component of the
Process Profile graphic is the Qualification Level, which is a formal
method of evaluating and ranking a process’ performance. The
Qualification Level, initially developed by IBM and later adapted to
healthcare by Shaw Resources, develops a quantifiable measure for the
"health" of the process so that everyone has an understanding
of how well that process is performing; the lower the number, the better
the performance.
The team started at level 6 – an
unknown level of performance. After six months, we moved to 4.08, into
the ‘Functional’ status of performance. Using the Process Advisor
software checklists, the Qualification score was automatically computed.
Measure the Process
As shown in Figure 3 above,
Marty and her team determined three types of measures:
-
Time-related: Time
from when a report is dictated until it is available. A three-month
analysis of turnaround times for selected types of reports – History
& Physicals, Operations, and Consultations – showed a stable
process with few deviations, and none of statistical relevance (see
Figure 4 below for an example chart). This same stability made it a
prime target for improvement, showing that the goal was easily
attained. Additionally, since the transcription process directly
affects one of the primary customers of the hospital (physicians and
other caregivers), any improvements will have a double impact by
increasing caregiver satisfaction.
Figure 4: Shows the
percentage of Consultation Reports turned around in 24 hours, which was
the standard
for turnaround time (TAT) for transcribed reports at the initiation of the
process improvement project.
-
Adverse
Indicators: # of incomplete reports; # of
reports that need QA; calls from caregivers inquiring about the report
status; increase in the # of delinquent records; # of reworked
reports; # of reports going to the wrong location; # of incomplete
reports filed in the chart.
-
Management
costs: No resources were added and no cost
increases were experienced outside those associated with increased
volumes. The team delivered improved quality of service with no
increase in unit cost – quality is free. Since the year-over-year
cost per line was the same, the cost was actually less after adjusting
for inflation.
Analyze the Process
Looking at all the baseline
measures for the time from when a report is dictated until it is available
(Figure 4), it was obviously a stable process with few adverse indicators
(reports delivered later than 24 hours after dictation). From here, the
team determined a new goal for the time-related measure, and wanted to
make it a stretch goal by cutting turnaround times by 50 percent: 12-hour
turnaround times for these reports.
Figure 5: This
Statistical Process Control Chart shows the number of statistically
significant defects
(outside three standard deviations from the norm) when the turnaround time
(TAT) is cut 50 percent -
from 24 hours to 12hours. (Compare to Figure 4.)
Investigation into the causes for
failing to achieve the 12 hour TAT uncovered a number of interesting
sources contributing to the variation:
Figure 6: Shows the
number of reports dictated each week over 24 hours for a period of three
weeks.
Colored boxes show times covered by medical transcriptionists Red boxes
delineate reports
dictated outside the covered hours before scheduling changes were made.
24/7 operation had
significant gaps in it – particularly Saturday night, Sunday,
midnights.
Weekend and evening
coverage was insufficient for volumes.
Inconsistency of
transcriptionists’ start times and hours worked per day, as shown in
Figure 7 below.
Figure 7: Original time
schedule for medical transcriptionists (numbers are start and stop
times)
shows considerable gaps in coverage during peak times of physician
dictation, delineated in red box
(see Figure 6).
Improve the Process
Initially, we identified
the most pressing ‘defect’ of the transcription process as reports not
printing in the correct place (going to the wrong unit/printer in the
hospital). We brainstormed the causes of this process deficiency. The
brainstorming revealed a surprising number of reasons behind this
shortcoming (see Figure 8).
Figure 8: This fishbone
diagram shows some of the various causes for a report printing in the
wrong location.
Armed with the findings from the
analysis, the team made process changes aimed at improving the measures of
timeliness, accuracy and printing to the correct location:
The following changes were
made on the dictation system in routing work to the transcriptionists:
-
Sequence of report
types was tailored for each individual pool. Formerly, Consults
gradually ‘fell to the bottom of the pool’ as new work types were
added, therefore taking longer to get done. Additionally, the number
of Consults were, at a minimum, two times greater than other reports
due to the number of specialists called in by the attending physician
for elderly patient care. Consults were therefore ‘realigned’ to
move up higher in the transcription pool.
-
Set up baseline
"aging" pools; this flagged older reports, allowing the
reports to be prioritized to appropriately meet the new 12-hour
turnaround time goal.
-
Developed
"Guidelines for Home-Based Transcriptionists;" these
guidelines were specific about the scheduling issues of times and days
that were not covered by a transcriptionist. Schedules were reworked
to make the turnaround time goals happen as required. Improved
communications for downtimes, time off, etc.
Information Systems-related
changes included–
-
Medical Record number
was entered twice, one with an extra "0" and without a room
number that caused the transcribed report to be sent to the wrong
location – this was corrected by IS.
-
Transfer patients
showed in two open units, causing the report to be directed to the
wrong location. This was fixed through programming changes.
-
One surprising
discovery in uncovering causes for reports printing in the wrong
location was that transcriptionists would select the wrong patient
type at the time of transcription. This error was eliminated by moving
the patient type field in the patient search screen to the top of the
screen. (See Figure 9)
Figure 9: Pareto chart
depicting transcriptionists selecting the wrong patient type -
five of the transcriptionists were involved in 80 percent of the errors.
Control the Process
The team went even further
in order to maintain control over the process and continuously improve it
through benchmarking. They surveyed caregivers and contacted other
hospitals, both internal to the BJC Healthcare system as well as external
organizations who were Baldrige National Quality Award Winners.
Figure 10: Results of the
treating physician survey.
Surveys tend to try to measure
satisfaction, which is a temporary state of mind, and therefore not a
preferred measure for quality or loyalty. So the Process Improvement Team
created a survey that instead measured how well the transcription services
met/failed to meet or exceeded treating physicians’ expectations.
The feedback from team members and
customers (caregivers) has been overwhelmingly positive. One physician
recently commented that "I dictated the report, and by the time I got
to my office, the report was there."
Additional control measures were put
in place, such as adding a Weekly Production Statistics Report to track
TAT, defects, etc, that helps the team manage the process more
effectively.
Benchmarks against other
organizations, especially against world-class Baldrige-winner hospitals,
gave the team a better understanding of where the Provide Transcription
Services Sub-Process stood. The team selected sister BJC Hospital Missouri
Baptist Medical Center as their internal comparison; and Baldrige winners
DePaul Health Center, St. Louis, MO, St. Luke’s, Kansas City, MO, and
Baptist Hospital, Pensacola, FL, as their external benchmarks.
Figure 11: This chart
shows how Christian Hospital’s Consultation Report transcription
services
Accomplishments
In a relatively short
amount of time – only six months from inception – the Process
Improvement Team has learned a great deal about the Provide Transcription
Services Sub-Process and made some exemplary improvements.
Figure 12: This
chart shows steady improvement in the number of overall guarantee failures
-
Manager (Marty) was set
up with a home-based workstation to monitor and address the escalation
process put in place (ie: fill in weekend slot, unusual number of
reports backing up in the queue, or a technical problem).
Daily emails on volume
status to Medical Transcriptionists increased to three times a day made
the status more real-time to provide transcriptionists with a better
view of volume.
Weekly Production
Statistics Report - new report internal to the team from Marty to track
TAT, defects, etc, to help manage the process more effectively.
Medical Transcriptionists’
Work Schedule is emailed three weeks in advance, which ensures
appropriate coverage 24/7. Transcriptionists previously did not see the
schedule at all and now schedules are sent out nine weeks in advance to
help all staff to organize vacations, and other personal time off.
Turnaround time
results/graphs discussed at monthly meetings so that everyone knows the
current status.
Transcriptionists
"stepped forward" to take hours to fill the early am and late
pm gaps – worked as a team to ensure that goals were met.
Rotation schedules
established for weekends – ensure coverage for all hours, and not
enough people who wanted to work every weekend – so started a rotation
schedule to give some people weekends off who needed/wanted them.
AMH Turnaround Time
improved as a byproduct – Alton Memorial Hospital is another
institution within BJC, and Christian Hospital does their transcription
as well.
Cross-trained
transcriptionists to do CH and AMH work types.
Overall physician
satisfaction with improved report delivery time has inceased. A
byproduct of the shorter report delivery times is that doctors remember
more of the patient interaction and are able to make better corrections,
thus, improving the quality of service to the patients.
Process Qualification
Level of 4.08 – This qualification level indicates that the process is
functional and all primary customer expectations are being met. The
process is systematically measured, and streamlining has begun.
Next Steps
In order to facilitate
continued improvements, the team identified several resources that would
be required: the availability of a data resource person; capital resources
to implement speech recognition and E-sign capabilities for physicians to
electronically sign dictated reports; and expansion of the reporting
capabilities of source system. Future objectives and challenges have been
ascertained as follows:
-
Implement new dictation
system 2006-2007
-
Speech Recognition
– This will provide additional convenience for physicians; the
transcriptionists will become more like editors to improve
productivity; cost savings and productivity to be determined.
-
National average
age of Medical Transcriptionist = 52(average age at Christian
Hospital is 50); speech recognition will help prevent repetitive
injuries (carpal tunnel, etc,) that affect older workers.
Additionally, the new technology may attract younger workers; a
major issue is that fewer people are selecting transcription as a
career option.
-
Increase / decrease
volumes - no prediction available – either increase or decrease can
be challenging with staffing:
-
Transcribed Lines
increased 2.9% 2005
-
2006 so far is up
9.6% – continued measurement will determine whether it is a
trend, or just a seasonal flux.
-
Provide Electronic
Signature - by second quarter 2007; this was found to be the ‘end of
process’ from the treating physician’s point of view, while the
transcription department considered it as a separate process.
Providing this capability will decrease the time for paperwork for the
physicians; also appeals to physicians who are technologically savvy,
thus helping to recruit new admitting physicians.
-
Provide alternate
report distribution methods, i.e. email (currently using autofax,
inter-office mail, US mail, pick-up) – another satisfier for
physicians’ offices.
-
Seek feedback/survey
from customers – this is part of the continued measurement,
analysis, improvement cycles of the Customer-Inspired methodology.
-
Communicate service
guarantee by end of Q2 2006
-
Process Qualification
Level
Summary
Dramatic, quantifiable
results have been achieved within six months. Christian Hospital has not
only cut report turnaround times from 24 hours to 12 hours or less, but
has increased physician satisfaction regarding timely reports.
Shorter turnaround times
have enabled dictating physicians to now focus on the quality of the
reports. Since they receive the report in less time after speaking with
the patient, they remember more of the patient interaction and are able to
make better corrections, thus, improving the quality of service to the
patients.
All the quality
improvements have come at little or no cost, reinforcing the adage that
quality is free. The benefits will be adding up for years to come as the
same process improvement methodology that the team learned for the
transcription process will also serve them well in dealing with future
technologies.
The process improvement
would not have been started as soon as it did without the buy in and
support of the CFO. It is paramount that senior executives lead the charge
for process improvement within the organization, and support
recommendations for future improvements. With the backing of senior
management, results like the ones realized by the Transcription Department
can be pushed throughout the organization, one department at a time.
Copyright
© Shaw Resources, 2006, all rights reserved. (888-742-9737), email: Info@ShawResources.com;
www.ShawResources.com. You may
reproduce this article provided: 1) each copy you generate is of the
article in its entirety, without modification of any kind; 2) you receive
no fee whatsoever; and 3) this copyright and permission notice, including
the contact information, must be prominently displayed on each copy
produced. |
