Reducing Patient Harm
by Improving Medication Turn-Around Time at a Children's
Hospital
In late 2002, we examined occurrence report
data to determine how to reduce medication errors. Though
delay of medications initially was not thought to be a significant
error type, our data indicated otherwise. Prior years' aggregate
analysis of incident reports showed 9% of medication error
reports resulted from delay. Approximately 23% of errors
involving medications resulting in patient harm were due
to delay. Incident reports documented nursing frustration,
parent dissatisfaction and patient harm.
In early 2003, we formed a team of nurses,
pharmacists, pharmacy technicians and unit-coordinators.
We defined and measured turn-around time (TAT) for medications
supplied by the main pharmacy, not in floor stock. TAT begins
from the time the physician placed the new medication order
to the time the medication left the pharmacy. Observational
measurements were taken through sampling during the quarter.
Responses from list-serve questionnaires helped determine
our goal of 85% of medications turned around in <90 minutes
with a stretch goal of 95%.
Our baseline data found a mean TAT was 87
minutes with a high standard deviation, measured through
sampling data. Pharmacy process flow was inefficient. Pharmacist
order entry workload was unevenly distributed. Cart fill
was inaccurate and required significant rework. Intravenous
and oral syringe medications were made during the time when
most discharge or discontinue orders were being processed
resulting in wasted intravenous medications accounted for
nearly $490,000 per year.
We used lean processing principles through
a rapid improvement workshop split into a 3-day design workshop,
followed by a 2-day implementation-planning workshop six-weeks
later. We designed a new pharmacy workflow by eliminating
many non-value added steps. The pharmacy model moved a large
portion of the manufacturing process to evenings, eliminating
a lot of rework. We created a new streamlined-order processing
staffing model without added FTEs. One pharmacist on day
and evening shifts sole responsibility is new-order checking
and sending. Medication rooms, with a low patient to medication
room-ratios, were designed to replace cluster-desk medication
carts to reduce RN search time, as well as rework in the
pharmacy. Our design also encompassed process flow eliminating
wait times when computerized physician order entry (CPOE)
came into place.
Following implementation of new pharmacy model
in summer 2003, our median turn-around time was reduced
from 75 minutes to 52 minutes. The following quarter, our
institution implemented CPOE, and our median turn-around
time was reduced to 23 minutes, a 69% improvement from baseline!
Currently, 97% of medications are turned around in less
than 90 minutes, up from 66% at baseline.
In the three months following implementation,
there were no incident reports related to delay in medication
TAT resulting in patient harm. Additionally, only 2.6% of
incident reports written in this time period were due to
medication TAT delay, down from 9%.
As of now, medication rooms are in place in
three wings of our hospital with plans to retrofit five
more wings before year-end. We continue to collect data
to assure our improvements are sustained. The lean processing
and rapid improvement cycle continues to be a successful
change methodology within our organization.
We used this high profile project as an example
of the importance of completing incident reports, and how
incident reports are aggregated to generate patent safety
improvements.