Injury and illness associated with drugs are major problems in Australia and around the world, despite significant developments in the area of adverse drug reaction (ADR) decision support technology. The aims of this thesis are: to investigate the ADR decision domain; to determine factors that may assist in the prevention, detection and management of ADRs; and, to inform the pre-requirements analysis phase of the development of decision support systems. An approach has been taken that permits open and grounded study of the decision environment. This approach can then be used to frame and inform the design of an ADR decision support system. Fifteen case studies that comprise self selected consumers, the treating medical practitioner/s and expert perspectives of a single instance of an ADR (fifteen in-depth consumer interviews, eight in-depth medical practitioner interviews and 30 expert written questionnaires), have been collected and analysed using a grounded theory approach, a symbolic interactionist theoretical framework and a social constructionist epistemology. The analysis was performed from three perspectives: individual case study analysis (all interviews for an instance of an ADR); group analysis (consumer, medical practitioner and expert views) and analysis combining the individual case studies and groups of data. Concepts, themes and theory have emerged from these data in the following areas: • the contribution of the differences in understanding of the core concepts within this domain, to misunderstandings between decision-makers; • the consumer as a diagnostic decision-maker in the ADR decision domain; • differential diagnostic strategies used by the consumers and medical practitioners; • complexities in the ADR decision domain that make diagnosis difficult; • the role of ADR information in consumer and medical practitioner decision-making; • decision types used by consumers and medical practitioners in the ADR decision domain; • resources used by consumers, medical practitioners and experts to inform their ADR decisions; • decision-making with partial knowledge of the consumer case history, drug behaviour and diseases; • the impact of suspected ADRs on consumers and on future decision-making; • medical practitioner/consumer decision-making models; and, • reasons for low ADR reporting and the impact on the development of new ADR knowledge. The results above suggest the following: • The ADR decision domain is more complex than the current ADR decision support focus and that broadening this focus may assist in providing a more complete and useful decision support solution. • Improving the prevention, detection and management of ADRs requires more than providing prescribers with up to date ADR information. Other important factors are sharing of information, awareness of the role of the consumer, a collaborative approach between the consumers and medical practitioners, and generation of new ADR knowledge. • A grounded theory analysis of case study data using the theoretical perspectives of social constructionism and symbolic interactionism provided insight into this domain from the perspectives of multiple decision-makers. This may be an approach that can be used by systems analysts to inform the requirements analysis phases of decision support within other domains. The results of this qualitative work are preliminary. Future work is required to confirm and expand these results. / Doctor of Philosophy

Injury and illness associated with drugs are major problems in Australia and around the world, despite significant developments in the area of adverse drug reaction (ADR) decision support technology. The aims of this thesis are: to investigate the ADR decision domain; to determine factors that may assist in the prevention, detection and management of ADRs; and, to inform the pre-requirements analysis phase of the development of decision support systems. An approach has been taken that permits open and grounded study of the decision environment. This approach can then be used to frame and inform the design of an ADR decision support system. Fifteen case studies that comprise self selected consumers, the treating medical practitioner/s and expert perspectives of a single instance of an ADR (fifteen in-depth consumer interviews, eight in-depth medical practitioner interviews and 30 expert written questionnaires), have been collected and analysed using a grounded theory approach, a symbolic interactionist theoretical framework and a social constructionist epistemology. The analysis was performed from three perspectives: individual case study analysis (all interviews for an instance of an ADR); group analysis (consumer, medical practitioner and expert views) and analysis combining the individual case studies and groups of data. Concepts, themes and theory have emerged from these data in the following areas: • the contribution of the differences in understanding of the core concepts within this domain, to misunderstandings between decision-makers; • the consumer as a diagnostic decision-maker in the ADR decision domain; • differential diagnostic strategies used by the consumers and medical practitioners; • complexities in the ADR decision domain that make diagnosis difficult; • the role of ADR information in consumer and medical practitioner decision-making; • decision types used by consumers and medical practitioners in the ADR decision domain; • resources used by consumers, medical practitioners and experts to inform their ADR decisions; • decision-making with partial knowledge of the consumer case history, drug behaviour and diseases; • the impact of suspected ADRs on consumers and on future decision-making; • medical practitioner/consumer decision-making models; and, • reasons for low ADR reporting and the impact on the development of new ADR knowledge. The results above suggest the following: • The ADR decision domain is more complex than the current ADR decision support focus and that broadening this focus may assist in providing a more complete and useful decision support solution. • Improving the prevention, detection and management of ADRs requires more than providing prescribers with up to date ADR information. Other important factors are sharing of information, awareness of the role of the consumer, a collaborative approach between the consumers and medical practitioners, and generation of new ADR knowledge. • A grounded theory analysis of case study data using the theoretical perspectives of social constructionism and symbolic interactionism provided insight into this domain from the perspectives of multiple decision-makers. This may be an approach that can be used by systems analysts to inform the requirements analysis phases of decision support within other domains. The results of this qualitative work are preliminary. Future work is required to confirm and expand these results. / Doctor of Philosophy