Statement of the problem: The risk-benefit profile of warfarin anticoagulation in hemodialysis (HD) patients differs compared to the non-HD population. Computerized decision support systems (CDSS) to assist with anticoagulation management are safe and effective in the non-HD population but had not previously been studied in HD outpatients. Methods of investigation: A before – after study compared anticoagulation control during pre-existing, nephrologist-led anticoagulation management to that following implementation of a pharmacist-led, CDSS-assisted strategy, in HD patients on warfarin at The Ottawa Hospital. Results: Forty-two patients were included. Following implementation of the CDSS-assisted strategy, median time-in-range increased by 3.7% (IQR, -9.5% - 20.6%; p = 0.247). Median frequency of INR tests per day decreased: -0.040 (IQR, -0.074 to –0.0008; P = 0.0001). Adverse events were similar. Conclusion: A CDSS-assisted strategy for anticoagulation management in HD patients is effective, safe and may lead to cost savings related to less frequent INR testing.

The overall aim of this thesis was to study the thinking strategies and clinical reasoning processes of registered nurses (RNs) and to implement and test a computerized decision support system (CDSS) integrated into the electronic health care record (EHR) to improve patient outcomes, i.e. to prevent pressure ulcers (PUs) and malnutrition among residents in nursing homes.  A think-aloud (TA) study with a purposeful sample of RNs (n=30) was conducted to explore their thinking strategies and clinical reasoning (Paper I). A quasi-experimental study with a convenience sample of residents (at baseline, n=491 and at follow-up, n=480) from nursing homes (n=15) allocated into two intervention groups and one control group was carried out in 2007 and 2009 (Paper II). In Paper III residents’ records were reviewed with three instruments. Nursing personnel (n=25) from four nursing homes that had used the CDSS for eight months were interviewed and the CDSS was tested by nursing personnel (n=5) in two usability evaluations (Paper IV). The results showed that the RNs used a variety of thinking strategies and a lack of systematic risk assessment was identified (Paper I). The proportion of malnourished residents decreased significantly in one of the intervention groups after implementing the CDSS, however there were no differences between the groups (Paper II). The CDSS resulted in more complete and comprehensive documentation of PUs and malnutrition (Paper III). The nursing personnel considered ease of use, usefulness and a supportive work environment as the main facilitators of CDSS use in nursing homes. Barriers were lack of training, resistance to using computers and limited integration of the CDSS within the EHR system (Paper IV). In conclusion, the findings support integrating CDSSs into the EHR in nursing homes to support the nursing personnel.

Medication management constitutes a large part of registered nurses' (RNs) daily work in municipal elderly care. They are responsible for monitoring multimorbid older persons with extensive treatments, and they often work alone, without daily access to physicians. RNs’ drug monitoring is, in this thesis, based on the concept of pharmacovigilance. Pharmacovigilance is about the science and the activities that aim to improve patient care and safety in drug use, that is, to detect, assess, understand and prevent drug-related problems. The overall aim was to explore conditions for pharmacovigilance from a nursing perspective, focusing on implications of RNs’ competence and use of a computerized decision support system (CDSS). Both quantitative and qualitative research methods were used, including a questionnaire (I), focus group discussions (II), individual interviews (III) and an intervention study (IV). In total 216 RNs and 54 older persons participated from 13 special accommodations, located in three different regions. RNs who had completed further training in pharmacovigilance rated their medication competence higher than those who had not. However, there was no difference between groups in the number of pharmacovigilant activities they performed in clinical practice (I). The RNs appeared to act as “vigilant intermediaries” in drug treatment. They depended on the nursing staff's observations of drug-related problems. The RNs continuously controlled the work of staff and physicians, and attempted to compensate for shortcomings in competence, accessibility and continuity (II). RNs’ use of a CDSS was found to affect drug monitoring, including aspects of time, responsibility, standardization of the work, as well as access to knowledge and opportunities for evidence-based care (III). The CDSS detected significantly more drug-related problems when conducting medication reviews, than the RNs did. Nevertheless, this did not result in any significant improvement in the quality of drug use in the follow up, three and six months later (IV). This thesis contributes to the recognition of pharmacovigilance from a nursing perspective. Increased medication competence seems to be insufficient to generate pharmacovigilant activities. RNs depend on other health care professionals and organizational conditions in order to perform their work. A CDSS has the potential to support RNs, both in structured medication reviews and in daily clinical practice. Inter-professional collaboration is crucial, with or without a CDSS, and the entire team needs to be aware of and take responsibility. Other important conditions is the existence of well-functioning communication channels, competence across the team, and established procedures based on current guidelines.