A fuzzy risk assessment model for hospital information system implementation

• Published in: Expert systems with applications 2012, Vol.39 (1), p.1211-1218
• Main Authors: Yucel, Gulcin, Cebi, Selcuk, Hoege, Bo, Ozok, Ahmet F.
• Format: Article
• Language: English
• Subjects: Analytic network process; Decision making; Fuzzy; Fuzzy inference system; Fuzzy set theory; Health care; Hospital information system; Hospitals; Information systems; Mathematical models; Risk; Risk assessment
• ISSN: 0957-4174; 1873-6793
• DOI: 10.1016/j.eswa.2011.07.129

► In this study, a fuzzy risk assessment model for Hospital Information System (HIS) implementation is improved and applied in a Turkish Hospital. ► The model processes experts’ evaluations defined in linguistic forms when there is no sufficient data. ► The model consists of analytic network process (ANP), reality-design gap evaluation and fuzzy inference system. ► The model reveals the relative importance’s of risk factors and design-reality gaps. ► The model can be used for risk prediction and risk management. There is research which reveals negative effects of IT applications in the healthcare sector on both patients and staff. Therefore, methods are necessary to predict the risk of new healthcare information technology in order to reduce the unintended results of new applications. A new predictive risk assessment model for a hospital information system (HIS) has been developed in this paper to estimate risk before the implementation of new HIS. The methodology consists of analytic network process (ANP), reality-design gap evaluation and fuzzy inference system. An application of the proposed algorithm has been applied for a research and education hospital in Istanbul, Turkey. Risk magnitude of a new HIS implementation for the hospital is found as major with a belief of 100%. The relative importances of risk factors for HIS implementation success are obtained. The most effective factors on the HIS implementation are found as technological factors; usefulness, compatibility, user involvement and ease of use. These factors are followed by organizational factors; training and organizational commitment. The most important individual factor is also found as user’s previous HIS experience. A risk assessment model has been proposed in this paper. The model processes experts’ evaluations defined in linguistic forms when there is no sufficient data and it integrates possible risk factors into the decision-making process of risk assessment. In the model, a reality-design gap analysis is used to determine risk likelihood instead of directly risk evaluation.