Beruflich Dokumente
Kultur Dokumente
note
R
ising prescription drug use, a
growing elderly population, and Purpose. The results of a workflow analysis The theoretical application of two mod-
reduced reimbursement rates at a large central outpatient pharmacy are els of enhanced workflow indicated that
reported, with theoretical modeling of model A (expansion of pharmacy techni-
have pressured the pharmacy profes-
potential efficiencies attainable through cians’ standard duties to include prescrip-
sion to meet societal prescription workflow enhancements. tion preparation, stock-container retrieval,
needs by creating more comprehen- Summary. In keeping with concepts of and prescription processing) could reduce
sive and efficient pharmacy services. “lean health care,” a time–motion analysis pharmacists’ non-value-added time by
As demand for prescription medica- was conducted at a central outpatient about 55%, or more than six minutes per
tions continues to grow, improving pharmacy that dispenses an average of prescription, with an even greater (74%)
workflow management in outpatient 250 prescriptions per day. Through direct potential reduction offered by model B
observation over an eight-week period, (technician performance of checkout pro-
pharmacy settings (i.e., community
pharmacists’ dispensing-oriented activities cedures in addition to expanded standard
pharmacies and hospital outpatient were categorized as either value-added duties). Although the research site was
pharmacies) is vital to providing (i.e., centered on direct pharmacist–patient atypical in its high staffing level relative to
a higher quality of patient care. 1 contact and, hence, providing increased prescription volume, the findings suggest
From 1994 to 2009, the number of value to the patient) or non-value-added. that similar workflow enhancements might
prescription drugs dispensed annu- The workflow analysis suggested oppor- be applicable in a range of community
ally in the United States increased by tunities to derive more value from phar- practice settings.
macists’ time by shifting their efforts away Conclusion. Through analysis of existing
86% (from 2.1 billion to 3.9 billion)
from non-value-added activities (i.e., tech- workflow in an outpatient pharmacy, op-
while the U.S. population grew just nical dispensing functions) toward value- portunities to optimize the use of value-
14%.2-4 Adapting to the pressures of added activities: engagement of patients added pharmacist time in the dispensing
a more complex health care system on entry into the pharmacy, pharmacist process were identified.
may require fundamental changes to order verification, and patient counseling. Am J Health-Syst Pharm. 2012; 69:966-71
drug distribution models, including
reducing the role of pharmacists in
technical dispensing activities and
emphasizing increased pharmacist– cists more effectively use their time pharmacists were spending just 31%
patient interactions.5,6 for direct patient interaction rather of their time performing clinical ac-
Research has shown a need for than technical dispensing activities. tivities associated with drug therapy,
improved workflow in outpatient According to the results of an inde- whereas over 60% of their time was
pharmacy settings to help pharma- pendent study published in 1999, being spent performing tasks that
Alex Jenkins, Pharm.D., M.S., is Manager, Ambulatory Pharmacy Lauren McKnight, Pharm.D., CPP, and Keith Overfield are
Services, WakeMed Health & Hospitals, Raleigh, NC; at the time acknowledged for contributions to the research described in this
of writing, he was Resident, University of North Carolina (UNC) article. The staff of the central outpatient pharmacy at UNC
Hospitals and Clinics, Chapel Hill. Stephen F. Eckel, Pharm.D., Hospitals and Clinics is acknowledged for participating in and
M.H.A., BCPS, FAPhA, FASHP, is Assistant Director of Pharmacy, providing feedback on this research.
UNC Hospitals and Clinics, and Clinical Assistant Professor, The authors have declared no potential conflicts of interest.
Eshelman School of Pharmacy, UNC, Chapel Hill.
Address correspondence to Dr. Eckel at the Department of Copyright © 2012, American Society of Health-System
Pharmacy, University of North Carolina Hospitals and Clinics, 101 Pharmacists, Inc. All rights reserved. 1079-2082/12/0601-0966$06.00.
Manning Drive, Chapel Hill, NC 27514. DOI 10.2146/ajhp110389
could be transferred to ancillary Hospitals and Clinics (UNCH) cen- over an eight-week period (March–
personnel.7 Results of a survey con- tral outpatient pharmacy is designed May 2010) in the UNCH central
ducted in 2000 indicated that outpa- to provide a hig h deg ree of outpatient pharmacy. All informa-
tient pharmacists were devoting only pharmacist–patient interaction, but tion was collected by direct observa-
28% of their time to patient-directed the allocation of pharmacists’ time tion and elicitation of staff feedback
interactions, with 56% of their time has not been optimized for patient in order to accurately evaluate the
devoted to dispensing tasks.8 Redi- care due to pharmacists’ involve- unique workflow of the outpatient
recting pharmacists to use more of ment in technical dispensing tasks. pharmacy. Observations were con-
their time for patient interaction The outpatient pharmacy, which ducted on weekdays from 8 a.m. to
requires modifications of pharmacy dispenses approximately 240 pre- 5 p.m. by the study investigators and
workflow that facilitate a redistribu- scriptions daily, is staffed by up to six two research assistants.
tion of dispensing activities. pharmacists and six pharmacy tech- Pharmacist activities. For the
A growing body of evidence in- nicians at any given time. The unusu- first seven weeks of the analysis, qual-
dicates that workflow enhancements ally high number of staff relative to itative observations were collected
that shift technical dispensing activi- the prescription volume is intended periodically by one investigator (ap-
ties away from the pharmacist (i.e., to make pharmacists available for proximately 15 hours per week spent
the use of automation and increased direct patient interactions such as in the pharmacy) in order to describe
technician-focused dispensing re- patient profile review, medication and differentiate the pharmacist ac-
sponsibilities) could be used as ef- reconciliation, and patient counsel- tivities that characterized the outpa-
fective tools to increase pharmacist– ing. The outpatient pharmacy is tient pharmacy workflow. Examples
patient interactions. 9-11 While the also equipped with a queuing sys- of qualitative observations included
benefits of enhanced workflow pat- tem (Q-Matic, Qmatic United States, determining the level of pharmacist
terns have been demonstrated in the Fletcher, NC) and an automated involvement in a certain activity and
literature, there is little information dispensing system (PharmASSIST, deciding whether or not that involve-
that delineates a process for conduct- Innovation, Johnson City, NY) to ment ultimately added value to the
ing a comprehensive evaluation of streamline medication dispensation. patient. During the final week of the
the use of pharmacist time during However, despite having a staffing analysis, workflow measurements for
medication dispensation. model and automated solutions to a time–motion analysis were collect-
One way to perform a detailed facilitate a high level of pharmacist– ed by research assistants (a total of
characterization of pharmacist time patient interaction, patients still wait approximately 40 hours spent in the
is to identify the “value-added” time to be seen because pharmacists are pharmacy). Because staff feedback
(i.e., time that is used to add value for slowed by the dispensing functions was a vital component of this project,
the patient) in the dispensing process. that technicians could be performing. the pharmacists and other pharmacy
That can be achieved by applying the Performing a detailed evaluation staff were apprised of the study ob-
concepts of “value stream mapping”— of the UNCH central outpatient jectives. The research was approved
an offshoot of “lean manufacturing” pharmacy’s workflow was critical to by the University of North Carolina
practices pioneered by car makers and identifying opportunities for opti- institutional review board.
now widely used in health care and mizing the application of pharma- Workflow analysis. The out-
other fields—to analyze the workflow cists’ skills in order to reduce patient patient pharmacy workflow was
involved in bringing a product to the wait times and improve the overall analyzed in order to measure phar-
customer or patient.12-16 quality of patient care. macist workload by describing the
The value that pharmacists pro- The management of workflow in fundamental “activity sets” involved
vide to patients is derived from the the UNCH central outpatient phar- in medication dispensing and then
use of their expertise to promote macy was analyzed by evaluating examining the amount of pharmacist
safe medication use through patient the use of pharmacist time during time dedicated to each activity set
interactions. This article describes medication dispensation; as a result, throughout the dispensing process
a workflow analysis and a proposed opportunities for improvement that (Table 1). For the purposes of this
methodology for conducting a robust could be used to develop future plans analysis, an activity set was defined
workflow evaluation with the ultimate for workflow enhancement were as a distinct activity or group of ac-
goal of allowing pharmacists more identified. tivities in the dispensing process that
time for direct patient interactions. had defined start and end times. The
Evaluation of existing workflow degree of pharmacist involvement
Background This prospective, observational in each activity set was determined
The University of North Carolina workflow analysis was conducted by observing whether activities were
Table 1.
Results of Workflow Analysis at UNCH Central Outpatient Pharmacya
Pharmacist Service
Activity Set Description Involvement Type
Queue time Wait time before pharmacist engagement NA NA
Pharmacist engagement Profile review, medication reconciliation, Yes Value-added
prescription entry and adjudication
Prescription preparation Retrieval of prescription labels 50% Non-value-added
Stock retrieval Retrieval of medication stock containers 50% Non-value-added
Prescription processing Scanning, counting, pouring, labeling 50% Non-value-added
Pharmacist verification Product verification Yes Value-added
Will-call preparation Collation and preparation for will call or Yes Non-value-added
patient counseling
Return to stock Stock-container return No NA
Counseling–checkout Patient counseling and checkout Yes Partial value-addedb
UNCH = University of North Carolina Hospitals and Clinics, NA = not applicable.
a
Patient counseling.
b
the time patients entered the outpa- Theoretical modeling. Using the ment of at least 50%, but only three
tient pharmacy. calculated baseline figures, theo- entailed value-added work. Because
All data in this assessment were retical models were developed to the initial patient engagement was al-
collected using handheld devices demonstrate opportunities for im- ways performed by a pharmacist, the
equipped with UMT Plus software proved use of pharmacist time by same pharmacist often performed
(Laubrass, Inc., Montreal, Canada), showing that the application of dif- all subsequent activity sets in the
which is designed to capture data ferent workflow enhancements could dispensing process (except return to
for work measurement studies. Data minimize pharmacist involvement in stock). This resulted in pharmacists
collection was based on repeated non-value-added activities, thereby using approximately 43% of their
observations of the performance facilitating increased pharmacist– time for non-value-added work
of each activity set. Because some patient interactions. Two models instead of directing more attention
activity sets involved multiple pre- were developed: toward expediting queue times by
scriptions and significant variability initiating engagement with patients
was anticipated between time mea- • Theoretical model A—Expansion of earlier.
surements, all measurements were pharmacy technicians’ defined re- Because pharmacist–patient in-
standardized by prescription volume sponsibilities to include three techni- teractions were strongly emphasized
(i.e., the process time for each mea- cal dispensing activities (prescription in the baseline dispensing process, a
sured activity set was divided by the preparation, stock-bottle retrieval, significant amount of value-added
corresponding prescription volume). and prescription processing) pharmacist work was involved, but
The calculated standardized times • Theoretical model B—Technician there were still opportunities to im-
were then represented as the aver- performance of checkout procedures prove the use of pharmacist time.
age pharmacist-dedicated time per (after patient counseling by a phar- Theoretical modeling demonstrat-
prescription. macist) in addition to the above-listed ed how non-value-added pharmacist
The mean dedicated pharmacist expanded standard duties work could be reduced by applying
time for each activity set was deter- technician-driven workflow enhance-
mined by multiplying the average Calculations resulting from the ments: an expansion of technicians’
dedicated time per prescription by theoretical application of models A defined responsibilities (model A) and
the percentage of pharmacist in- and B indicated that the proposed technician performance of checkout
volvement. Calculated figures for workflow enhancements would activities in addition to an expansion
dedicated pharmacist time were then greatly reduce the amount of non- of defined responsibilities (model B).
used to define a baseline model of value-added pharmacist time dedi- The theoretical application of these
how pharmacist time was used dur- cated to the processing of dispensed workflow enhancements showed that
ing medication dispensation in the prescriptions. With models A and non-value-added pharmacist time
outpatient pharmacy. B, the total value-added pharmacist could be reduced, thereby reducing
During the data collection period, time would remain the same as at the pharmacists’ share of the total
the overall dispensing process time baseline (8.41 minutes per prescrip- process time, by shifting technical dis-
was 40.74 minutes per prescription tion), but non-value-added pharma- pensing activities from pharmacists to
(that included a queue time of 21.9 cist time would be reduced to 2.83 pharmacy technicians.
minutes before pharmacist engage- and 1.66 minutes per prescription, If broadly implemented, measures
ment), and the mean total measured respectively (Table 2); as a result, to reduce the amount of pharmacist
time was 18.84 minutes per prescrip- the total pharmacist process time time dedicated to technical dispens-
tion (Table 2). Because pharmacists per prescription dispensed would be ing activities would facilitate a reallo-
were highly involved with dispensing reduced by 24% under model A and cation of the pharmacist’s work load
activities, pharmacist time repre- 32% under model B. to enable an increased focus on direct
sented a significant portion (14.73 patient interactions in any outpatient
minutes per prescription) of the Discussion or community pharmacy setting.
total measured time, and pharmacy At baseline, the workflow in the The potential benefits of optimizing
technician work accounted for just UNCH central outpatient pharmacy pharmacist–patient interactions in-
4.11 minutes per prescription. Fur- was characterized by significant clude improved workflow efficiency,
thermore, approximately 43% of pharmacist involvement in technical increased opportunities for MTM,
pharmacist time (6.32 minutes per aspects of the medication dispens- a reduced salary cost per dispensed
prescription) dedicated to dispens- ing process. Of the nine activity sets prescription, an enhanced ability to
ing was determined to be non-value- identified in the workflow analysis, handle a larger prescription volume,
added time. seven entailed pharmacist involve- a reduced risk of medication errors
Table 2.
Results of Time–Motion Analysis of Dispensing Process at UNCH Central Outpatient Pharmacya
Mean ± S.D.
Total Time per Mean ± S.D. Pharmacist Time per Prescription (min)
Variable Prescription (min) Baseline Model Ab Model Bc
Activity sets
Pharmacist engagement 5.08 ± 4.72 5.08 ± 4.72 5.08 ± 4.72 5.08 ± 4.72
Prescription preparation 1.42e 0.71e . . .d ...
Stock retrieval 1.10 ± 1.56 0.55 ± 0.78 ... ...
Prescription processing 4.45 ± 4.44 2.23 ± 2.22 ... ...
Pharmacist verification 2.15 ± 1.71 2.15 ± 1.71 2.15 ± 1.71 2.15 ± 1.71
Will-call preparation 1.66 ± 0.74 1.66 ± 0.74 1.66 ± 0.74 1.66 ± 0.74
Return to stock 0.63 ± 1.45 ... ... ...
Counseling–checkout 2.35 ± 2.99 2.35 ± 2.99 2.35 ± 2.99 1.18 ± 1.50
Time calculations
Total prescription process time 18.84 ± 17.61 14.73 11.24 10.07 ± 8.67
Value-added pharmacist time ... 8.41 8.41 8.41 ± 7.93
Non-value-added pharmacist time ... 6.32 2.83 1.66 ± 0.74
a
UNCH = University of North Carolina Hospitals and Clinics.
b
Expansion of pharmacy technicians’ defined responsibilities to include prescription preparation, stock retrieval, and prescription processing.
c
Pharmacy technician performance of checkout activities in addition to expanded defined responsibilities per model A.
d
Not evaluated.
e
S.D. not calculable.
due to process standardization, and, tion, the results of the observational and workflow evaluations at other
ultimately, improved quality of pa- analysis described here would have outpatient pharmacy practice sites
tient care. had greater external validity if the are warranted.
The workflow analysis and theo- theoretical workflow enhancements
retical modeling exercise had several had been implemented during the Conclusion
notable limitations. First, the patient study period; if that had been the Through analysis of existing
care model at the UNCH central case, preimplementation and post- workflow in an outpatient pharmacy,
outpatient pharmacy is atypical in implementation data collection opportunities to optimize the use of
that the number of pharmacists on might have allowed a stronger dem- value-added pharmacist time in the
staff is high relative to the prescrip- onstration of the potential benefits dispensing process were identified.
tion volume; while most outpatient of adopting workflow modifications
pharmacies that dispense around 250 designed to optimize pharmacist– References
prescriptions per day are likely not patient interactions. 1. American Pharmaceutical Associa-
tion, National Association of Chain
staffed by more than two pharma- Despite those limitations, com- Drug Stores, National Community
cists at any given time, the UNCH bining time–motion studies with Pharmacists Association. Implement-
outpatient pharmacy is simultane- concepts of lean health care pro- ing effective change in meeting the
demands of community pharmacy prac-
ously staffed by up to six pharmacists vided a robust understanding of tice in the United States (August 1999).
in order to emphasize direct patient how pharmacist time could be used www.pharmacist.com/AM/Template.
interactions. Second, throughout more effectively in the UNCH central cfm?Section=Home&TEMPLATE=/
CM/ContentDisplay.cfm&
the workflow evaluation, there was outpatient pharmacy. Regardless of CONTENTID=2576 (accessed 2012 Jan
significant variability in the observed the outpatient pharmacy’s atypical 4).
pharmacist and technician activities staffing model, the basic elements of 2. Kaiser Family Foundation. Prescription
drug trends (May 2010). www.kff.org/
(i.e., rather than completing each ac- the pharmacist workflow evaluation rxdrugs/upload/3057-08.pdf (accessed
tivity set without interruption, they can be applied to any outpatient or 2012 Jan 4).
often stopped performing one activ- community pharmacy setting. Iden- 3. Kaiser Family Foundation. Prescription
drug trends (June 2006). www.kff.org/
ity set to perform other, unrelated tifying opportunities for improved rxdrugs/upload/3057-05.pdf (accessed
activities); because of that measure- workflow may also be useful in ana- 2012 Jan 4).
ment variability, almost 50% of the lyzing strategies to reduce medica- 4. U.S. Census Bureau. Population dis-
tribution and change: 2000 to 2010.
initially collected data were excluded tion errors by introducing standard- www.census.gov/prod/cen2010/briefs/
from the workflow analysis. In addi- ized processes. Similar time–motion c2010br-01.pdf (accessed 2012 Jan 4).