Improving Strategic Management of Hospitals: Addressing
Functional Interdependencies within Medical Care
Dr. Masanori Akiyama
Daniel Goldsmith
Dr. Michael Siegel
Sloan School of Management
Massachusetts Institute of Technology
Center for Digital Business, Room NE20-336
Cambridge, MA 02142
Voice: (617) 253-2937
Fax: (617) 258-7579
{poas, goldsmith, msiegel} @ mit.edu
Abstract
To better understand the performance of hospital operations in response to IT-enabled
improvement, we report the results of a system dynamics model designed to improve core
medical processes. Utilizing system dynamics modeling and emerging HIS data, we
demonstrate how current behavior within the hospital leads to a ‘stove-pipe’ effect, in
which each functional group employs policies that are rational at the group level, but
that lead to inefficiencies at the hospital level. We recommend management
improvements in both materials and staff utilization to address the stove-pipe effect, and
estimate the resultant cost-saving. We believe that the major gains in health information
systems use will accompany new information gathering capabilities, as these capabilities
result in collections of data that can be used to greatly improve patient safety, hospital
operations, and medical decision support.
Introduction
This work is designed to better understand how hospitals respond and adapt to the
introduction of Health Information Systems (HIS). The current work extends previous
hospital research on nurses and the injection process to include a wider set of
stakeholders. This work examines the entire supply chain of patient injections with
consideration ward, pharmacy and doctor impacts on operations. The motivation for this
research stems from the observation that while the need for new HIS in hospital
environments has been well documented, system managers, as well as medical
practitioners, have both recorded their disappointment with many HIS implementations
(Kohn, 2000).
Part of the reason for suboptimal performance is that many approaches to HIS fail to take
full advantage of the new opportunities provided by data collections systems as a tool to:
a) understand, measure, and track hospital operations, b) identify and implement high-
leverage improvements, and c) provide opportunities for hospital staff to train and learn
more effectively. Our research suggests avenues to utilize the rich data set provided by
HIS to improve hospital efficiency, patient safety, and the receptiveness of staff to IT
enabled-improvements. Comprehensive HIS allow for the development of new tools to
support effective integration and success operation of new systems.
This research draws from the analysis of the HIS in place at several major Japanese
hospitals, named POAS (Point of Act System). As described by Akiyama (2001, 2007),
the underlying concept of POAS is to enable records of “‘who did what to whom, where,
when, using what, and for what reason.’ In short, real-time input becomes possible at the
point of action.” Under the system, logs of medical actions and inventories are created
throughout the course of operations, recorded using bar-code scanning technology and
nurses equipped with PDAS (personal digital assistants). The system operates
continuously at the hospital, handling 100 transactions per second, or more than 360,000
transactions per hour, and has been in continuous operation for four years. For example,
the system collects information on every interaction between order, drug, nurse, and
patient. Utilizing this data, we can revisit the challenges associated with HIS and
understand system wide behavior.
As described by Akiyama, Siegel, and Goldsmith (2007), soon after implementation,
POAS facilitated vast improvement in multiple areas of hospital operations, with
estimated savings reaching millions of dollars. In addition to POAS-enabled cost savings,
the system also led to impressive improvement in patient safety. Prior to the
implementation of POAS, there existed nearly a 40 percent chance that there would be a
misadministration of an injection prescription, due to the absence of an automated
method of checking injections and the lack of real-time communication. After POAS, this
percentage was cut dramatically; an alarm would sound prior to the injection if any
problems existed (such as a correct patient being presented with an incorrect medication),
and the staff would be able to fix the mistake prior to injection. In the years following the
initial implementation of POAS, patient-safety benefits continued to be realized, and by
all measures, improvement remained robust. However, concern was raised about the
sustainability of the system’s financial performance. System managers were concerned
about how to obtain further improvements in of the hospital’s financial performance. Of
particular concem were the areas of overlap between functional groups within the
hospital. For example, for a patient to receive an injection, information and materials
must be effectively shared by doctors, pharmacists, and nurses. The ability for system
managers to help manage these interactions was thought to be a key determinant to
overall system efficiency, as measured in staff and materials utilization. Further, POAS
improvements had created a rich-data set of operational data that was being underutilized
in hospital improvement. We combine analysis of POAS data sets with site visits to
POAS hospitals, systems dynamics modeling, and expert input to derive important
process changes across the hospital system.
The Silo Effect in Hospital Ward Management
The basic injection process provides a useful way of thinking about the challenges
associated with hospital ward management. The injection process at POAS hospitals,
also described in Akiyama, Siegel, and Goldsmith (2007), refers to the different paths an
injection order can take, culminating with either in a successful injection or in a changed
or cancelled order.
The key actors in the injection process are doctors, pharmacists, and nurses, and Figure 1
presents a ward view of operations. This figure is a conceptual simplification of actual
hospital operations, but highlights each actor’s key procedures. The normal path of
injection goes from left to right: a doctor issues a prescription order, the pharmacist
package the set of drugs required (referred to as an Rp) and checks the order for
correctness then the nursing staff mixes the Rp components together and injects it into
the patient. The solid arrows below each actor’s name signal the observables for each
actor: for example, pharmacists rarely have any information about the downstream
processes of the nursing staff. Their viewpoint is restricted to their core operations,
preparing and checking orders. In other words, the current injection process creates a
“stovepipe effect,” by which each key actor operates within a functional silo. While
hospital managers may view the entire process, doctors, nurses, and pharmacists are all
bounded in visibility by the specific breadth of their function.
Also of note in Figure 1 are the flows of information and materials, as represented by the
dashed arrows. Information and materials can flow downstream—as in the normal
injection process described above—but it can also flow upstream, as in the case of
changed orders. In upstream operations, a nurse will check with the doctors to see if the
order has been changed prior to injection. If the order change is processed prior to the
mixing phase, the Rp components can be returned to stock by the pharmacists and are
generally reusable for future patient orders. It is important to note that if the Rp has
already been mixed prior to the change order, the change must result in the disposal of
the Rp.
Doctor's Pharmacist’s Nurse's
View View View
2S et en ----
) 5
iet---/4@ 2 F ff
airs > \---4 9
Manager's 95 g a o
i a5 8 z 9 3
View z 8 a ~
Bis) |-----| [§ é |___l § v
3 a
eee eae —
Figure 1. Differing Views of Hospital Operations
The difference between orders that can be reused and those that must be thrown out has
been shown to cause significant variations in the amount the hospital spends on drug
inventory, as well as the efficiency with which the staff processes an order. In both
instances change orders result in inefficiencies in the system. The “Manager's View” sees
the effects across the silos and can better understand these “global” inefficiencies.
In this research, we highlight the central role of pharmacists in efficient management of
the injection process, and examine the ability for pharmacists to mange operations given
their limited visibility of overall operations. We show that pharmacists have developed
self-defeating policies as a result of the stove-pipe effects that maximize downstream
flows while ignoring the consequences of upstream flows. We then calculate the costs of
these policies, and recommend a series of policy interventions to ameliorate the negative
impacts.
Pharmacy Operations
In the section below we present the modeling formulations of ward operations that have
emerged from our research. The theory reported here draws on extant literature in process
improvement, system dynamics modeling, as well as our study of one hospital setting
(Forrester, 1958; Repenning & Sterman, 2002). In particular, we unpack pharmacy
operations as a balancing process, by which pharmacists must balance competing
demands for their time. Ultimately, we see that pharmacists pursue strategies that they
erroneously believe reduce their work burden and lead to waste in medication. While the
consequences of pharmacist’s actions are difficult to see from their perspective, given
their limited visibility of other operations, and lead to significant downstream
inefficiency and may even impact patient safety.
Despite these negative consequences, we demonstrate that the current strategies
pharmacists employ to manage their work load are intendedly rational. The problem is
that they are developed from a flawed mental model of hospital operations. Instead of
faulting pharmacists for the negative outcomes we observed, we show how structural
limitations in hospital operations create an environment in which well intentioned polices
at the group level (i.e. pharmacists) can subvert overall operational efficiency. To arrive
at this insight, we present a series of causal diagrams to capture hospital dynamics,
construct a simulation model to quantify system behavior over time, and simulate the
model, presenting the results from the perspective of multiple hospital actors.
The core pharmacy operations are represented in the balancing loop B1 in Figure 2. The
rate of new orders from doctors increases the work that the pharmacists must complete.
Depending on the delay in pharmacy, the orders will be processed at a given rate: the
shorter the delay, all else equal, the rate of operations will increase. The result of
pharmacy operations, preparing and checking drugs, here referred to as filling orders,
closes a loop by diminishing the amount of work left to do. The “work accomplishment”
structure in B1 captures the embedded rationality of pharmacists: their goal is to reduce
the amount of work left to accomplishment.
Pharmacy Delay
+
Work to DOF rom 2
+ New Orders Implied Rate of
Operations
New Order Rate - 4er)
Filling Orders
Figure 2. Intended Rational of Pharmacy Operations
Figure 3 presents a constraint on the rate of hospital operations: the capacity of pharmacy,
a concept that combines staff and equipment into an overall metric of how fast, at its
maximum, the pharmacy would be able to process orders. First, the capacity of the
pharmacy is represented as a fundamental constraint on operations, shown in loop B2.
The capacity of the pharmacy sets a feasible rate of operations, given an amount of work
to do, representing the upper limit of pharmacy operations. The capacity of the pharmacy
also has a role in determining the normal pharmacy delay: the time in which, on average,
it takes the pharmacy to fill an order. While not the only influence of pharmacy delay
(which include such other factors as work habits and expectations), the capacity of the
pharmacy influences the pharmacy delay (another way to think of the pharmacy delay is
the average time an order will spend within the purview of the pharmacy.)
Capacity of
Pharmacy Normal
Pharmacy Delay
fa
Feasible Rate of +
Operations Pharmacy Delay
wehia
>
* New Orders Implied Rate A
Operations
New Order Rate - -)
@s
Filling Orders
Figure 3. Capacity Constraints
To examine these basic pharmacy dynamics in the context of a sample hospital, we first
consider data from POAS on the changes in work burden on pharmacists over time. For
several reasons, including structural changes in the Japanese healthcare environment,
both the number of patients and the number of orders rose over a period of several years.
(Figure 4) Orders, however, rose faster than patients, and the increase created a rise in
the orders per patient. During the same time the overall staff in the pharmacy was
constant. Therefore, pharmacists were faced with increasing work burdens.
Orders/Patient
2
15
uw
105 /~
10 oo
95
9
85
8
2002 2003 2004 2005 2006
Year
Figure 4. Changes Order per Patient Over Time
As a result of the changes in the new order rate over time, the pressure on pharmacists
increases. We have captured the concept of work pressure, which can be thought of a
ratio of implied rate of operations divided by the feasible rate of operations. That is, the
closer a pharmacist gets to the maximum feasible rate he or she can work, the more
pressure they feel.
Capacity of
Pharmacy ional
Pharmacy Delay
Feasible Rate of —~ +
Operations a Pharmacy Delay
Work to Do From PS
+ New Orders Implied Rate of +
Operations
New Order Rate ‘i BL
Gs
Filling Orders
Figure 5. Work Pressure
We represent the rise in orders shown in Figure 6 below with an arrow, and have circled
several possible options available to managers and pharmacists to mange the increased
work burden. The three options are to: a) increase the capacity of the pharmacy, by, for
example, hiring more staff; b) lengthen the normal pharmacy delay, likely requiring
patients to wait longer for orders to be complete; and c) increase the implied rate of
operations by finding some way to work faster with the same capacity.
Capacity of
Pharmacy onal
[Pharmacy Delay
Hire More
YS Wait Longer
Work Pressure
Feasible Rate of For Ordersg
Operations Pharmacy Delay
. S
‘Work to Do From
#
+ New Orders Implied Rate of
Operations
New Order Rate .
Work Faster
i] iif
Figure 6. Options to Manage Increases in Work Pressure
While all the options conceptually reduce work pressure, both hiring more staff and
requiring a longer wait for orders are likely difficult to accomplish and may be, in fact,
infeasible. Increasing the work rate, therefore, is the most appealing option. Further, it is
under control of the pharmacists themselves, and allows them regulate work pressure
without the intervention of other actors. While managers must become involved to hire
more staff, and nurses and patients will be affected by longer delays, pharmacists, by
developing a way to work faster, can balance the work burden with little perceived costs.
One way to accomplish this is to complete orders in batches, an operation represented by
loop B3. (Figure 7)
Capacity of
Phamacy Normal
Pharmacy Delay
Batch
eo
— via
Feasible Rate of
Operations Phamaty Delay
Work to Drom 2
+ New Orders Implied Rate of
Operations
New Order Rate : Aer)
@s
Filling Orders
Figure 7. Batching
Batch processing combines orders into batches which are processed in delivered in large
groups and allow pharmacists to increase the rate at which orders are processed.
While new orders represent the majority of the work pharmacists must accomplish, order
changes also creates work for pharmacists to accomplish. The rate of order changes also
creates another stock of work for pharmacists to accomplish. The demands from order
changes, however, creates competing demands, represent by loops B4 and R1. Working
to fill order changes reduces the pile of order change, but reduces the rate at which new
orders can be processed. From interviews with pharmacists conducted at a sample
hospital, the pharmacists were clear that they considered new orders the most important
to fill, and the backlog of new orders to be the most salient indicator of how effective
they were in managing their work load. Pharmacists would likely pursue policies to favor
loop R1 over loop B4. (Figure 8)
Capacity of
Pharmacy Neiat
Pharmacy Delay
Batch
Operations
ws
Work Pressure Fe
4
Feasible Rate of Ap
Operations i: Pharmacy Delay
Work to D&*F rom
fi New Orders implied Rate of
Operators “okt Do Fram
4e1) )
An)
New Order Rate
&
Filling Orders
Work to oe)
Changes
+
Rate of Order
Changes
Figure 8. Balancing Work Pressures
Also, that batching operations in loop B3 also have a virtuous property of reducing work
from redos, as shown in loop B5 of Figure 9. Batching reduces the average time the
order stays in the pharmacy by speeding up the rate at which they fill orders. As orders
spend less time in the pharmacy, fewer changes are, on average, made while the order is
at the pharmacy, reducing the amount of work pharmacists must do from order changes
in the pharmacy. Batching, therefore, appears to be not only an effective strategy to
manage new orders, it also reduces the amount of work from redos. A pharmacist
commented on this, effect saying “We would like to get redos in the pharmacy down to
zero.” (Figure 9)
Capacity of———
Pharmacy Normal
Pharmacy Delay
Batch
peti
— ZZ
wat fore
~ +
Feasible Rate of haa
‘Operations id a Pharmacy Delay
Work to Dd From
+ New Orders imple ae
Operations
tier 20 From
Key Orde ae
Work to wv
Changes
New Order Rate i)
(@f
Filling Orders
Rate of Order
- charaes
85h order changes in °
4. Pharmacy
4
= Time for Order
Number of Orders Changes in __
with Nurses Pharmacy
Figure 9. Reducing Changes at Pharmacy
Based on the dynamics demonstrated in the causal diagrams above, we have constructed
a simulation model (shown its entirety in Appendix 1).We first simulated the model from
the pharmacist’s perspective for a one-week period, and present the results in time series
graphs in Figure 10. We see the positive benefits on batching from the pharmacy
perspective. The base case (no batching) is shown in solid lines, and the batching case is
shown in dashed lines. As a result of batching, work to do goes down, as the completion
rate goes up, and the numbers of redos that occur in the pharmacy also fall.
Work to Do in Pharmacy Completion Rate Redos in Pharmacy
Figure 10. Simulation of Batching in Pharmacy
Conceptually the mental model that leads to the batching strategy is limited. The results
presented in figure x are skewed by this partial view of possible outcomes, rather than a
whole-systems perspective. Figure 11 explicitly includes feedback from the nurses, which
is not represented in the pharmacist’s conceptual model of operations. Loops R2a and
R2b show the key feedback from redos at the nursing station; in both loops the redos that
the pharmacists believe they are avoiding by batching reappear. Pushing orders faster out
of the pharmacy does reduce redos orders while they are in the pharmacy, but only by
delaying the same redos until the nurses are working on them. Loop R2a shows this
feedback, with the key delay for these orders to return to the pharmacy explicitly
recognized by the double line across the causal link. In other words, the redos from the
nurses are the very same redos the pharmacy would have processed if the order had
stayed in the pharmacy longer (even though pharmacists don’t see this.)
Loop R2b shows how this feedback is observed from the pharmacy perspective. The
redos from the pharmacy are not treated as redo orders; instead, when they appear from
the nurses, they seem to be new orders. Pharmacists, because of how the orders are
processed, lose the traceability of ordering that would let them match up the subsequent
order change with the new order. The perceptual differences between loops R2a and R2b,
which correspond to different mental models of pharmacy operations, are crucial to
understanding pharmacy behavior.
Capacity Of
Pharmacy Sorat
Pharmacy Delay
Batch
Operations
7
Work Pressure ,
Feasible Rate of “9 af
Operations —~ Pharmacy Delay
Number of Work wo —~
i ork to Do From
Patients y New Orders Implied Rate a
Operations “orc to Do From
New Order Rate . Order Changes
yA +y
Filling Order
a ow
Work to wd
Change Orders Changes
Generated From Nurses +
(Pharmacy View) ity
o @} Order Changes in Rate of Order
Number of Orders Pharmacy Changes
with Nurses = Time for Order a
Changes in oS Change Orders
pon Pharmacy Generated From
Nurses
+
Average Duration of
+
Order with Nurses omarctaiees
in Ward
Figure 11. Batching Reduces Changes at Pharmacy
The Role of Pharmacy Operations on Downstream Processes
We now include nurses into our causal framework and observe the downstream
consequences of batching. To do so, we return to the possible outcomes of an injection
order; it can be a) successfully injected, b) changed before mixing while at the pharmacy,
c) changed before mixing while at the nurse station, or d) changed after mixing at the
bedside. There are important differences between outcomes: most importantly, once the
order is mixed, it must be wasted, as it can’t be used for other patients. Additionally, an
order returned from the nurse station has accumulated more time being processed than
one changed at the pharmacy. These differences are demonstrated in Figure 12, a stock
and flow structure that captures the four different outcomes (shown as outflows from
each stock.) The first series of constructs relate to the flow of orders and material. The
flows, denoted by straight arrows with values, are the rate at which orders for injections
are successfully moved (referred to as Rps, a bundled collection of injection materials)
between stations in the hospital. Figure 12 also shows three of the system’s stocks,
denoted by a rectangle, which are computed as the integration of the stock’s inflows less
its outflows. The stocks are the accumulation of orders and waiting to be processed at
three stages, the pharmacy, the nurse station, and the patient's bedside.
injections Rps at
a Nurse ead ie z
RpOrder [Pharmacy] pharmacy Station Mixing jedside | injecting into a
Rate Processing Rate Patient
Returning to Stock Returning to Stock Wasting
(From Pharmacy) (From Nurse Station)
Figure 12. Batching Reduces Changes at Pharmacy
In Figure 13 below, the consequence of batching are captured with arrows indicating the
implied changes in the variables; beginning with the pharmacy processing rate, as the rate
increases, injections in the pharmacy decrease, as do returns from the pharmacy. At the
same time, Rps at the nurse station rise (given a constant mixing rate). The dashed
rectangle shows the limits of the pharmacy’ perspective.
popu
I ti i Rps at
Ly Nurse we] Rest a
Rp Order | Pharmacy] Pharmacy Station Mixing Bedside | injecting into a
Rate | Processing Rate Patient
Returning to Stock | Returning to Stock Wasting
(From Nurse Station)
(From Pharmacy) |
Figure 13. Batching Reduces Changes at Pharmacy
Next, we capture the perspective of the nurses in Figure 14, and again show the implied
behavior with arrows. It has been demonstrated in previous research that nurses are likely
to respond to rising orders at the nurse station also with a batching strategy.'
| The insights are shown are also reflected as additions to the causal diagram in A ppendix 2.
Injecti | [7 Rps at
ee Je} Nurse Rps at
RpOrder [Pharmacy] pharmacy | |_Station | mixing Bedside | injecting into a
Rate Processing Rafe Patient
Returning to Stock Returning to Stock Wasting t
(From Pharmacy) ; (From Nurse Station)
Figure 14. Batching Reduces Changes at Pharmacy
As intended under POAS, each Rp is scheduled in advance to be mixed at a specific time,
which corresponds to the staggered schedule of each injection. However, nurses often
mix Rps in large batches throughout the day, which clusters the nurse’s workload and
increases the blocks of available downtime. This balances the increase in the pharmacy
processing rate, and accelerates the rate at which Rps are moved from the nurse station to
the bedside (i.e., in the mixed form). This however, leads more change orders to occur
while the orders are mixed, and these orders must be thrown out.
Simulating our model from the nurses perspective, we see a very different outcome as a
result of pharmacy batching, as shown in Figure 15. The base case is the solid red line,
and the batching case is shown in solid blue line. .Instead of the improvements we
witnessed from the pharmacy perspective, pharmacy batching has lead to an increase in
the nurse work rate, an increase in the number of orders wasted, and lead to an overall
increase in the time that staff must spend on order changes.
Nurses Work Rate Changes (Waste) with Nurses Total Time Spent on Changes
Figure 15. Simulation results from the nurse’s perspective
We support our simulations by using POAS data, and examine the behavior our
simulation suggests for a one-week period over a period of years. The amount of orders
that are either wasted or returned increases after initial gains were made in from 2002 to
2003. The percentage of orders wasted rose from 1.2 percent in October 2003, to 1.4
percent in October 2006. Additionally, the number of orders returned from nurses
increased from 8.9 percent to 10.4 percent over the same period. (Figure 16)
Number of Orders
16000
14000
12000 aos
10000 — a Injected
8000 —e Wasted
6000 Returned
4000
2000 — o
0 L L L
2002 2003 2004 2005 2006
Figure 16. Injection Outcomes
POAS data also verifies the relationship demonstrated in the simulation results shown
above. By relating the mixing gap, the time between when an order is mixed and when it
is injected, to the percentage of orders wasted, we see a positive relationship between
increase in the mixing gap and increases in the percentage of orders that are wasted.
(Figure 17)
% Waste 200
1.80
1.60
1.40
1.20
1.00 {a
0.80 |
B Ff EEE
0.40
0.20 = = | |
0.00
< (- 1-60 G1- 121- 181- 241- 301- 361- 421- 600<
500 500)- 120 180 240 300 360 420 600
0 Mixing Gap (In minutes)
Figure 17. Mixing Gap and Percentage Waste
In addition, POAS data shows a negative mixing gap for a small segment of orders. That
is, the orders are shown to be mixed after they have already been injected—a practical
impossibility. This data suggests evidence for an additional phenomenon in nurse
behavior that may result from pharmacy batching: early injection. Because nurses: a)
have the orders earlier, and b) are mixing them earlier, there is evidence to suggest that
nurses are injected them prior to the scheduled injection time. Some of these orders are
then changed, meaning patients received the wrong injection. Examining POAS data we
see that orders are injected early nearly 3.3 percent of the time. While the consequences
of early injection require further investigation, the evidence suggests it is at least possible
that pharmacy processes are having unintended downstream consequences on patient
safety. Further, because of the stove-pipe effect, pharmacists, without policy intervention,
would fail to realize the magnitude of consequences of their actions.
Recommendations
Our analysis supports a recommendation to not only reduce batching in the pharmacy, but
to actually increase the pharmacy delay for filling orders. Figure 18 shows three
simulation results: a) the base case (solid line); b) the batching case shown above (dashed
line); and c) the increased pharmacy delay case (dashed-dot line).In the increase
pharmacy delay case, orders spend, on average 50 percent longer in the pharmacy, but
doing so creates a meaningful reductions in cumulative work to do for order changes and
order retums for nurses (as more changes are caught in the pharmacy before additional
nurse work must be done.)
Further, we see opportunities to utilize data in pharmacy operations. For example, POAS
data indicates that a small set of medications disproportionately account for waste. Using
feedback from data to schedule operations (e.g., filling order, missing), and increasing the
delay in the pharmacy for these select medications, would produce significant savings in
materials and staff utilization. For example, by having the pharmacists hold the top 5
drugs, which account for nearly 25 percent of waste, as long as possible and avoid early
mixing by the nurses, we estimate potential savings of approximately 70 million yen, or
600 thousand US dollars, on an annual basis. Further improvements could be made by
dynamically linking operational data to pharmacy operations.
a Cumulative Time Spent on Order
Orders Pending in Pharmacy Changes Returns from Nurses
Figure 18. Increasing Pharmacy Delay Improves Metrics
Conclusion and Discussion
The goal of research in this area is to develop management improvements by means of
systems modeling and analysis combined with the use of newly available operational data
sources, such as POAS. We expect that this style of research will help sustain and advance
system-wide improvements in operational efficiency in a hospital setting. Ongoing
investigations will focus on designing robust experiments and simulations to increase
understanding of the causes and effects for sustained improvement. While implementation
challenges to this process improvement exist, we have proposed a high-leverage yet non-
disruptive improvement to quickly demonstrate financial benefits. Continued studies,
including additional staff interviews, will focus on better understanding the culture involved
in ward management to develop a workable solution for more comprehensive improvements
in ward management.
Utilizing system dynamics modeling and emerging HIS data, we have demonstrated how
current behavior within the hospital leads to a ‘stove-pipe’ effect, in which each
functional group employs policies that are intendedly rational at the group level, but that
also lead to inefficiency across operations at the hospital level. Our data suggests that
critical determinants of success in efficient hospital operations include the perceptions
stakeholders have about the effects of the actions on upstream and downstream processes.
Faulty attributions about the drivers of efficiency can trap operations in deteriorating
modes of performance, and subvert the momentum gained from IT-enabled processes.
We recommend management improvements in both materials and staff utilization to
address the stove-pipe effect, and estimate the resultant cost-saving. As part of this
analysis we also demonstrate opportunities to merge real-time operational data with
feedback modeling to provide dynamic tools for hospital administration, risk
management, and education and training. We believe that the major gains in health
information systems use will accompany new information gathering capabilities, as these
capabilities result in collections of data that can be used to greatly improve patient safety,
hospital operations, and medical decision support.
In the future, we envision a system that merges newly available operational data sources (i.e.,
real-time POAS data), electronic medical records, and operational data into feedback models
that create dynamic ward management tools. We propose a system that would provide
information to hospital about efficiency metrics, and would provide managers a means to
change policies, such as which drugs to exempt from mixing, at any point during on-going
hospital operations. This platform would be open for improvements, promoting the
development of additional tools that improve operations and manage patient risk.
The data produced by the POAS system combined with computational social science
methodologies, particularly system dynamics modeling, is an exciting step forward in
understanding the dynamics of managerial improvement in hospitals. With the reported
initial results and continued research, we predict that this work will have a significant
impact on reducing hospital costs, improving patient safety, and accommodating
improvements in hospital staff operations.
References
Akiyama M, Kondo T.. Risk management and measuring productivity with POAS--point
of act system. Medinfo. 2007;12(Pt 1):208-12.
Akiyama M. Risk Management and Measuring Productivity with POAS - Point of Act
System. A Medical Information System as ERP (Enter-prise Resource Planning) for
Hospital Manage-ment. Methods Inf Med. 2007;46(6): 686-93.
Akiyama M., Migration of the Japanese health-care enterprise from a financial to
integrated management: strategy and architecture, Medinfo. 10(Pt 1):715-718,2001.
Linda T. Kohn, Janet M. Corrigan, and Molla S. Donaldson, Editors. To Err Is Human:
Building a Safer Health System Committee on Quality of Health Care in America,
Institute of Medicine. 2000.
Repenning NP, Sterman JD. Capability Traps and Self-Confirming Attribution Errors in
the Dynamics of Process Improvement. Administrative Science Quarterly 2002;
47(2):265-295.
Appendix 1. Simulation Model
x ‘Tine pType
esos ah Da
~~
Appendix 2. Full Causal Diagram
P Feasible Rate of, Batch
Operations Operations’
Capacity of
Pharmacy “4 7 Normal
Work Pressure 3 Pharmacy Delay
fa
? Actual
Wiedorram, Pharmacy Delay
amber of New Orders
atients, + +] #
a
pee as Work to Do From
4s) peas Order Changes
# +
New Order Rate 48) “phe
i * Filling Orders .
Change Orders
Generated From Nurses Work Fikonter
Renae ee) 9 Order Changes in
* «Pharmacy
Number of Orders etwas
a a Pharmacy Change Orde’s ate of Order
Ra Generated From Changes
Nurses
Average Duration of s +
Order with Nurses
pe Order Changes
inWard
‘Orders to Be’
Orders Already Retumed to
Propensity to Mixed by Nurses Pharmacy
Inject Early (Waste) ;
* +
Propensity to Waste
Total Work for Reuseable Orders
Nurses aa
Work Pressure
3 onNurses
Nurse Capacity
