Libsie, Mulugeta(PhD)Ayele, Tsegaye2018-06-262023-11-292018-06-262023-11-292008-10http://etd.aau.edu.et/handle/123456789/3422Medical Referral Decisions are clinical decisions by which clinicians determine referral indication, type of required services and selection of appropriate providers in medical referral systems. A referral indication is a decision made by clinicians to determine whether referral is needed or not for a patient case under consideration. These decisions are made in a clinical environment through referral systems that aim at providing of efficient healthcare services by improving patient outcomes and decreasing cost incurred and time spent for such services. The quality of referral decisions is highly dependent on the efficiency and soundness of the decision making process. This inherently complicated referral decision process depends on a complex mix of both clinical and non-clinical factors such as patient, clinicians and healthcare system determinants. Recently, a Multi- Agent Referral Decision Support (MARDS) framework [6] has been proposed with the aim of improving the quality of referral decisions. However, it doesn’t fully address the referral indication aspect that is a key component of the medical referral process, which may cause under-referral and over-referral problems. This thesis proposes a Multi-Agent Decision Support (MADS) model for Referral Indication, Service Identification and Local Consultation aspects of medical referral aimed at providing improved decision support to clinicians. The proposed decision support model undertakes the analysis of determinants related to the referral indication, service identification and local medical consultation. This aid is provided through the social agents that interact and cooperate in the clinical environment, which are designed to interact with the existing CIS (Clinical Information System) and the CKB (Clinical Knowledgebase) to fetch critical information which supports the analysis of decision making. It is believed that this model extends the MARDS framework by addressing the referral indication and service identification aspect for its realization. Moreover, the local medical consultation service is believed to address the communication and organizational challenges of the medical consultation process and in turn contributes for the minimization of over-referrals and helps to overcome clinical uncertainties. Keywords: Medical Referral Decisions, Referral Indication, Service Identification, Local Consultation, Multi-Agent Systems, Decision Support SystemsenMedical Referral DecisionsReferral IndicationService IdentificationLocal ConsultationMulti-Agent SystemsDecision Support SystemsThesis