J. Vander Sloten, P. Verdonck, M. Nyssen, J. Haueisen (Eds.): ECIFMBE 2008, IFMBE Proceedings 22, pp. 1581–1584, 2008 www.springerlink.com © Springer-Verlag Berlin Heidelberg 2009 Health Technology Management: Medical Equipment Classification N.F. Oshiyama 1 , A.C. Silveira 1 and J.W.M. Bassani 1,2 1 Universidade Estadual de Campinas (UNICAMP)/Departamento de Engenharia Biomédica, Faculdade de Engenharia Elétrica e Computação, Campinas, Brazil 2 Universidade Estadual de Campinas (UNICAMP)/Centro de Engenharia Biomédica, Campinas, Brazil Abstract — This paper presents two methods for medical equipment classification according to three maintenance indi- cators (NC C = annual number of corrective maintenances divided by the median of the number of corrective mainte- nances for the specific maintenance group; TM C = time spent on corrective maintenance per year divided by the median of the time for the specific maintenance group, $M C = annual cost of corrective maintenance divided by 6% of the depreciated acquisition cost of the equipment). The main tool was a simple database for each type of equipment. As shown in a previous paper, one of the methods is based on the fact that the indica- tor values increase as the equipment gets older. The other method is based on the ABC analysis, in which three limits were established based on Pareto’s law. Data of syringe infu- sion pumps from years 2004 - 2006 (database of the Center for Biomedical Engineering, University of Campinas) were used. One-way analysis of variance revealed that NCc, TMc and $Mc values increase significantly with age. Out of 50 syringe infusion pumps, 42% were classified as C, although 80% of the equipments were not in the 10-14 year category. The results of the ABC and age analyses were 96% coincident. Maintenance indicators seem to keep a relation with the equipment age and, if used in nationwide scale, they might be a robust tool for equipment classification. Unexpected behaviors may be rele- vant warnings to the Clinical Engineering staff, regulatory agencies and manufacturers interested in data mining. Keywords — Clinical Engineering, equipment maintenance, medical equipment, maintenance indicators, ranking I. INTRODUCTION Current technological development increases the variety of medical equipment available to the market. Since these new technologies usually create a disproportion between the costs and the resources available, Clinical Engineering (CE) work becomes more important and requires more efficient management methods [1][2]. One of the main problems in this effort, however, is to find the appropriate way to obtain quantitative information on equipment performance along its lifetime. This type of information may be useful to help regulatory agencies to identify aberrant behavior in medical devices and investigate its causes before patients are harmed or deficient results are delivered to particular health pro- vider. In practical terms, it is quite valuable for the CE group to know whether a specific device performs well according to some criteria that take into account a larger population of similar devices under care of different CE groups. This work proposes two methods for medical equipment classification based on three maintenance indicators, which could be used at any hospital, independently of the number of beds or pieces of equipment. One of the methods is based on the hypothesis that the maintenance indicators increase as equipment ages (Aging Method, AM). The second method is based on the ABC (“Always Better Control”) analysis and is presented as a simpler, more limited, but yet useful method (in short, ABCM). II. MATERIALS AND METHODS In order to calculate the three indicators in the AM, data were collected from the Center for Biomedical Engineering (CEB) of the University of Campinas (UNICAMP), Campi- nas, Brazil, database from 2004 to 2006. A Microsoft® Excel™ spreadsheet was prepared with medical equipment attributes, such as identification, manufacturer, model, serial number, location, acquisition price and date, as well as the number of corrective maintenances (NC), total maintenance cost ($M) and time (TM). The three annual indicators are then calculated as fol- lows: Year NC median NC NC L C ⎟ ⎟ ⎠ ⎞ ⎜ ⎜ ⎝ ⎛ = ) ( (1), where NC L is the number of corrective maintenances from the local maintenance group. Year TM median TM TM L C ⎟ ⎟ ⎠ ⎞ ⎜ ⎜ ⎝ ⎛ = ) ( (2), where TM L is the total corrective maintenance time from the local maintenance group. Year A M M D C ⎟ ⎟ ⎠ ⎞ ⎜ ⎜ ⎝ ⎛ ⋅ = $ 06 , 0 $ $ (3), where $A D is the depreciated acquisition cost.