The Hong Kong Polytechnic University (HONG KONG)
About this paper:
Appears in: INTED2009 Proceedings
Publication year: 2009
Pages: 2958-2965
ISBN: 978-84-612-7578-6
ISSN: 2340-1079
Conference name: 3rd International Technology, Education and Development Conference
Dates: 9-11 March, 2009
Location: Valencia, Spain
Prescribing medicines is a complex decision that a small change may result in serious consequences. Up to now, the therapy and medicines to be administered to a patient rely heavily on the expertise and knowledge of physicians. Many patients take multiple medicines thereby increasing the likelihood of prescription errors such as drug-drug interactions and side effects. This is particularly true for young physicians as they are typically inexperienced in understanding the clinical background of patients, they may not be able to prescribe the most appropriate medicine combination. Additionally, medicine information changes rapidly making it extremely challenging to learn from medical literary works.

Taking advantage of advanced information and communication technologies, numerous researchers propose the introduction of medical prescription education to young physicians through e-learning programs. Using secure communication network, the data stored in the electronic medical records can be shared within the entire organization and serve as an educational platform for young physicians to learn from the medical experts. However, the current e-learning program is static and cannot stay abreast with updated prescription information. The dynamic knowledge acquisition and modeling technique proposed in this paper would address this inadequacy.

In this paper, we present a dynamic Medical Prescription Education System (MEDPRES) designed to automatically acquire knowledge of effective prescriptions and model such prescription decisions through the use of statistical analysis. Each therapeutic result is represented as a medical case which encodes the patient’s clinical background information, symptoms, diagnosis and medicines prescribed. The collective knowledge encoded in all these cases are summarized in form of conditional probability that statistically ranks the effectiveness of specific medicines prescribed for particular clinical situations. The prescription results stored in the case base are acquired from the practice of a real medical group and validated by a panel of medical experts. MEDPRES will solicit inputs from the young physician through queries. These inputs are used to determine the best solution with explanation which can be compared to the physician’s original prescription. A system prototype is developed and tested by ten undergraduates in a Hong Kong medical school to evaluate the feasibility of such an approach.
e-learning, knowledge acquisition, knowledge modelling, medical prescription.