1 of 4

Cost Management of Medical Equipment Maintenance

L. S. Rocha1, J. W. M. Bassani1,2
1
Department of Biomedical Engineering, 2Center for Biomedical Engineering, Universidade Estadual de Campinas,
Campinas, Brazil

Abstract—In the present study we combine Activity how activities take place in the production of the service, to
Based Costing (ABC) with a microprocess-based custom- determine the support and service cost centers, as well as to
made management system used to control of the medical select and determine the cost drivers.
equipment maintenance service performed by a clinical The diagram in Fig. 1 is a simplified version of the
engineering group in a public health institution in Brazil. microprocess-controlled maintenance management system
Results show the cost of service orders calculated currently used in CEB. Each code, on arrows, refers to an
through the allocation of the expenditure per cost center specific microprocess that is activated according to the
to activities performed during the year 2003. As this status of the service order given in Table I. The service cost
model can estimate how the activities affect profitability, center pointed by the arrowhead is responsible for
managers can use ABC information to interpret possible performing the activated microprocess and setting the next
strategies needed to investigate the viability of cost status.
minimization.
AO, AR, EX
Keywords—Activity based costing, clinical engineering,
state-diagram, public health system
GEC
OU DP, VT
CO AO AE, AT
I. INTRODUCTION
EP, AV,
AE CO, OU
AE GRU EXE
The Center for Biomedical Engineering (CEB), CAC CAC
EE
manages maintenance of medical equipment of the public
OA AS AE, AT
health system of the Universidade Estadual de Campinas
AI
(UNICAMP), located in the state of São Paulo, Brazil. The CO
university’s medical area comprises a School of Medicine, a SUP
four hundred fifty bed general hospital, a one hundred
twenty bed women’s hospital, and about ten satellite
medical centers. The total equipment inventory includes SU
approximately 10,000 items with complexity ranging from Fig. 1. State-diagram of microprocess-based management of maintenance
simple sphygmomanometers to modern imaging systems. In at CEB - UNICAMP. See Table I to understand the meaning of each status.
the present study we aimed at combining Activity Based
TABLE I
Costing (ABC) [1] with our microprocess-based custom- MEANING OF EACH ALLOWABLE STATUS OF SERVICE ORDER
made management system [2] to investigate the viability of
cost minimization and possible money raise for maintenance Status Meaning
investments. OA Service order call
The procedures necessary to perform any service within
AE Place Service order on waiting list
an organization are designated activities [3]. ABC is a
technique applied to calculate costs, which relates the EE Perform the maintenance
activities performed in the so called cost centers to the EP Service done
services finally produced. The amount of money spent in CO Service order concluded
each cost center is linked to the activities through the cost AV EXE is waiting equipment to arrive in lab
driver, which is the function that selects activities within a
CG Equipment is under guarantee
subset of all activities and outputs the value of the function
correspondent to the sum of the time spent providing the VT Third party technician called for in site repair or advice
total number of activities performed in a service cost center. AI Acquisition of spare parts
The application of the ABC method depends on an AS Budget is authorized by SUP
infrastructure that systematically records data about the
AT GRU is waiting for authorization from SUP
routine of organization allowing identification of which and
AO GEC is waiting for budget
Note: This work is supported by Fundação de Amparo à Pesquisa do AR Equipment is waiting for removal by third party contractor
Estado de São Paulo – FAPESP, Brazil, under Grant # 01/05944-0.
 2 of 4

DP Proposal of inactivation of equipment can easily identify slow transitions associated to high costs.
OU Higher level administrative decision is required to proceed In table II, for example, SU-AE and SU-AR are new types
SU Purchase performed by the medical area unity of transitions to control purchase of expensive equipment
parts performed by the medical area unity. Notice in Table
EX Service order executed at third party companies
VIII that transitions of this nature are not very frequent,
certainly less than 2 % of total (not included in the 5 most
As the state-diagram provides a view of the dynamics of frequent types of service orders, Table VIII).
the medical equipment maintenance service focusing on the Another interesting point is that numbers in Table V
specific microprocesses we integrated ABC and data indicate that most of the technical work is performed in a
collected by the state-diagram aiming at designing a cost single day.
management model, which could help managers to evaluate Table VII highlights the sequence of transitions and the
the possibility of becoming competitive in this market niche. number of occurrences of the most frequent service orders in
2003, which correspond to approximately 78% of the
service orders produced in that year.
II. METHODOLOGY Table VIII shows the cost and the cycle time per service
order. Note that costs range from ~R$100.00 (US$33.00) to
The steps to design the ABC model to determine the ~R$150.00 (US$50.00) for approximately 70% of the
cost of service orders were: service orders.
• To calculate the yearly spending of support and service
cost centers [4]. We used a conventional spreadsheet to TABLE II
YEARLY DATA RELATED TO PROVIDER OF SUPPLIES (SUP)
estimate the three cost categories: salaries, depreciation and
general expenses per cost center. Average time
Transition Cost (R$)
• To allocate the cost accumulated in support cost centers to (day)
service cost centers [5], based on the number of workers per AI - AE 30.66 320.31
service cost center. AS - AE 19.00 198.50
• To identify activities and to calculate the value of function AS - EX 10.73 112.10
of their cost drivers. Here we established that each status
AS - SU 12.86 134.35
transition characterizes an activity. We calculated the value
of the function related to cost driver by the total time spent SU - AE 58.25 608.56
in a year to perform the number of occurrences SU - AR 78.75 822.73
correspondent to the status transitions performed in each Expenditure (R$) 261,423.07
service cost center [5].
Cost driver (day) 25,023
• To calculate the cost driver rate related to each service cost
center. We divided the cost of each service cost center by its Cost driver rate (R$/day) 10.45
respective cost driver.
• To calculate the cost of activities. We multiplied the TABLE III
YEARLY DATA RELATED TO ENGINEERING GROUP (GRU)
average time spent to perform each transition by the cost
driver rate related to the service cost center in which Average time
occurred the transition. Transition Cost (R$)
(day)
• To identify the types of service orders and calculate their AT - AE 4.23 18.13
cost. Each sequence of transitions from OA to CO AT - AR 5.73 24.56
characterizes the type of service order. For all transitions
AT - AS 1.10 4.72
included in each sequence, the average cost of service order
is determined by summing the cost of transitions [6]. In AT - AT 5.93 25.42
addition, the summation of the average time for transitions AE - EE 6.19 26.53
comprised in a sequence from AO to CO resulted in the Expenditure (R$) 305,776.54
cycle time per service order.
Cost driver (day) 71,331
Cost driver rate (R$/day) 4.29
III. RESULTS

Tables II to VI shows the expenditure per cost center1 as

well as the respective values of cost drivers and cost driver
rates calculated during the year 2003. From these tables, one
1
US$ 1.00 is approximately R$ 3.00 (Real is the currency in Brazil).
 3 of 4

TABLE IV DP - AE 16.29 179.91

YEARLY DATA RELATED TO CUSTOMER CARE CENTER (CAC)
DP - CO 10.65 117.62
Average time EX - AE 15.09 166.66
Transition Cost (R$)
(day)
EX - AR 13.68 151.08
AV - AE 4.68 31.39
AO - OU 12.29 135.73
CG - AE 34.26 229.83
AO - CO 28.00 309.24
CG - AO 28.00 187.83
Expenditure (R$) 239,138.06
CG - AR 27.84 186.76
Cost driver (day) 21,653
EP - AE 3.77 25.29
Cost driver rate (R$/day) 11.04
EP - CO 4.16 27.91
OU - AE 20.55 137.85
TABLE VII
OU - CO 10.00 67.08 MOST FREQUENT SERVICE ORDERS PRODUCED IN 2003
VT - AE 22.62 151.74
Service
Sequence of transitions Amount %
VT - AO 5.40 36.22 order
EP - EE 4.67 31.33 #1 OA→AE→EE→EP→CO 3,381 41.69
OA - AE 0.004 0.03
VT - CG 3.50 23.48 #2 OA→AE→EE→CO 2,278 28.09

Expenditure (R$) 231,429.11

#3 OA→AE→EE→AV→AE→EE→EP→CO 383 4.72
Cost driver (day) 34,499.07
Cost driver rate (R$/day) 6.71 #4 OA→AE→EE→AI→AE→EE→EP→CO 151 1.86

TABLE V #5 OA→AE→EE→DP→CO 150 1.85

YEARLY DATA RELATED TO TECHNICAL STAFF (EXE)

Average time TABLE VIII

Transition Cost (R$) YEARLY COST AND CYCLE TIME FOR SERVICE ORDERS
(day)
EE - AE 2.20 124.88
Transition #1 #2 #3 #4 #5
EE - AI 1.62 91.95
OA - AE 0.03 0.03 0.03 0.03 0.03
EE - AO 1.37 77.76
AE - EE 26.53 26.53 53.06 53.06 26.53
EE - AV 0.87 49.38
EE - EP 57.33 57.33 57.33
EE - CO 2.09 118.63
EE - CO 118.63
EE - DP 1.29 73.22
EE - AV 49.38
EE - EP 1.01 57.33
AV - AE 31.39
EE - VT 0.45 25.54
EP - CO 27.91 27.91 27.91
EE - CG 1.38 78.33
EE - AI 91.95
EE - OU 2.14 121.47
AI - AE 320.31
Expenditure (R$) 905,188.57
EE - DP 73.22
Cost driver (day) 15,947
DP - CO 117.62
Cost driver rate (R$/day) 56.76
Cost (R$) 111.80 145.19 219.10 550.59 217.40
TABLE VI Cycle time (day) 11.364 8.284 23.104 49.834 18.134
YEARLY DATA RELATED TO CLINICAL ENGINEERING MANAGER (GEC)

Average time
Transition Cost (R$)
(day) IV. DISCUSSION
AO - AE 3.50 38.65
AO - AO 2.00 22.09
As shown in the previous sections of this paper,
information produced by this methodology that
AO - AT 19.61 216.57
systematically integrates the concepts of state-diagram and
AR - AE 6.73 74.33
 4 of 4

ABC can be used to cost management of maintenance The increased demand for service orders # 1 and # 2
service carried out by the clinical engineering group. leads to an equivalent increase in the amount of transitions
It is important to realize that each microprocess that are required to produce them, resulting in reduction of
transition promptly points to specific areas involved in cost driver rates and consequently a decrease in the cost of
producing a particular type of activity. We noticed, for all types of service orders.
instance, that at least 5% (see Table VII, #3) of the service
orders involved equipment transportation to the CEB
laboratories. This is currently done by the health unit V. CONCLUSION
transport service. It is clear that optimization is required to
reduce cycle times. Since we know the specific equipment The methodology presented in this study evaluates the
involved in all trajectories, we are now implementing a cost of medical equipment maintenance performed by a
small lab in the general hospital to shorten distances and, clinical engineering group in a public health institution in
therefore save time and reduce costs. This has helped us to Brazil. As this model can estimate how the activities affect
stress to the administrative team in the medical area, the profitability, managers can use ABC information to interpret
importance of implementing these simple and necessary possible strategies to set competitive prices for specific
measures. service orders.
In 2003, we assessed the average hourly labor costs of
the third party engineering group representatives. Based on
this information, we found that the prices for service orders ACKNOWLEDGMENT
in medical equipment maintenance ranges from R$100.00 to
R$140.00 per hour plus costs of spare parts. We assumed We express our appreciation to Dr. Rosana Almada
our sampling is a realistic estimation of the market prices for Bassani for her helpful comments.
specific services. More than 90% of the companies are
located within 150 Km from our hospitals and therefore it is
possible to consider that at least 2 hours will be computed to REFERENCES
perform a simple 30 min to 1 h maintenance job. In this
case, a service order would easily cost in average R$240.00 [1] R. Cooper and R.S. Kaplan, “Measure costs right: Make the right
decisions,” Harvard Bus. Rev., pp. 96-103, Sep.-Oct. 1988.
(or US$80.00). This is more than most of our service orders
[2] J.W.M. Bassani, L.S. Rocha, M.L. Lüders and W.J. Bizinotto,
would cost (see Table VIII). If we sell services at this rate it “Micro-process based management of medical equipment
would not be impossible to set a small profit margin. maintenance,” in Proc. 2nd. Joint Conf. of the IEEE Engineering
However, we believe that at this rate small public hospitals in Medicine and Biology Society and the Biomedical
Engineering Society, EMBS2002BMES, Houston, TX, pp. 1942-
would not be able to pay for the service. That is probably
1943.
one of the reasons why most of them lack any type of [3] R. H. Ramsey IV, “Activity-based costing for hospitals,”
regular maintenance group or contract. Hospital & Health Serv. Admin., vol. 39, no. 3, pp. 385-396, Fall
Given the above scenario and the possibility of 1994.
[4] L.S. Rocha and J.W.M. Bassani, “Activity-based costing for
associating costs and activities we started to exercise
clinical engineering,” in IFMBE Proc. Series 2nd. Eur. Medical
strategies to reduce elapsed transition times specially those & Biological Eng. Conf., EMBEC’02, Vienna, Austria, vol. 3.
involving microprocesses such as AI, EX, AV, SU and a [5] L. Brem, “Cambridge Hospital Community Health Network –
few others also not included in the 70% most probable The Primary Care Unit” in Harvard Business School Case 9-100-
054, Harvard Business School Pub., Boston, MA, Feb. 2000.
service orders. This is possible if we consider that instead of
[6] J.A. Brimson, Activity Accoun.: An Activity-Based Costing
a R$240.00 rate we sell services at R$150,00 and avoid Approach. New York, NY: John Wiley & Sons, 1991.
trajectories #3 to #5 we may be closer to acceptable
contracts. Additionally, in order to provide maintenance
service to the small hospitals, CEB would need to redesign
the activities required to process a given service order
resulting in an increased amount of transitions, but the
above solution of selecting types of trajectories to offer has
proved successful.
An extensive preventive maintenance program is very
difficult to implement at UNICAMP since hospitals struggle
to equilibrate the number of required equipment to the
increased demand. CEB has progressively increased
preventive inspection inclusion plans to minimize this
problem. It is possible that the lack of preventive
maintenance is responsible for the current types of service
orders that require more complicated trajectories.