A review of the literature on blood stream infections caused by contaminated intravenous infusates which are prepared in clinical care settings found that this common nursing procedure poses at times a significant and life-threatening risk to patients. The guidance and regulations surrounding the preparation of intravenous drugs in clinical care settings suggests that this procedure is extremely complex and poses many different potential hazards to patients. This thesis set out to determine how the infection risks are being addressed in practice by asking the questions: ‘What is the system of intravenous drug preparation in clinical care settings in NHS Scotland?’ and, ‘How does it work in practice?’ Several data sources were utilised: six locations, in specialities where the literature identified significant outbreaks had occurred, were examined for potential contamination risk. Observations (78) of infusate preparations were undertaken and, where available, written procedures were compared with observed practices. Finally, analyses were made of 71 questionnaires, completed by the nurses who prepare intravenous drugs, regarding their opinions of the procedures’ safety and when they perform redundancy checks. The conclusion of this study is that the system of preparing intravenous drugs in clinical care settings by nurses is, as a consequence of potential infusate contamination, error-prone and unreliable. The reasons for this conclusion are now detailed. o Due to a lack of mandatory environmental standards, and the provision of poor environments, there is a risk of infusate contamination from environmental sources and consequently, a risk to patients of infusate-related blood stream infections (IR-BSI). o Some in use equipment poses contamination risks to patients’ infusates. Equipment that could reduce the contamination risk is not always available and in some instances such safety-enhancing equipment has been removed. o There are no complete written procedures which mirror what is done in practice. At present, from a human-factors perspective, it is not easy for the nurse to do the right thing, or to be sure exactly what is the right thing to do. o The procedure, in practice, has the required elements of an aseptic procedure, but the execution of the procedure is more often not performed aseptically. o The procedure of intravenous drug preparation as observed is mainly an interrupted aseptic procedure and as such the recommencement of the aseptic procedure requires repeated hand hygiene. o The nurses’ opinions of safety vary, as did their assessment of the infection risk to their patients, but it is clear that intravenous drug preparation is not a much-loved nursing procedure and some nurses find it very stressful. o There is no asepsis quality control built into the system. Aseptic steps are the least likely to be performed as a redundancy check compared to the mandatory checks of ‘right patient, right drug and right dose’. o The information available to the nurses, from the drug companies, from the makers of equipment and from national agencies does not identify with sufficient clarity the infection risks, or detail how to negate them. Suggestions for improvement to the six procedures and environments are clear once the procedure steps are colour-coded as either aseptic or non-aseptic; validity testing of these improvements is however, still needed. The systems’ vulnerabilities observed in this research appear to stem from a chain of external influences including an underestimation of the problem size and the actions needed to prevent it in evidence-based guidelines and mandatory guidance. This leads to poor recognition of the risk of IR-BSI in clinical practice. The problem of infusate contamination causing IR-BSIs is further compounded by the fact that it is not caused by a single organism and does not always present as a disease in real time, that is, over the lifetime of the infusion. As a consequence, this presents surveillance difficulties in terms of definitions, data collection and analysis. Finally, although the diagnosis of a blood stream infection for an individual patient remains relatively easy, it is not easy to recognise a contaminated infusate as the origin of the problem. All these challenges make both the recognition of the problem and agreement on prevention strategies, extremely challenging. In summary, the main conclusion of this thesis is that the preparation of infusates in clinical care settings, which occurs approximately 3,000,000 times a year in NHSScotland, is from an aseptic perspective, error-prone and unreliable. Recommendations to optimise patient safety include, changing the procedure locally and, with the utmost urgency, the production of minimum environmental standards. The results of this study are relevant to all hospitals in Scotland and throughout the United Kingdom where the current regulations apply and similar procedures are performed.