Discussion

This in-depth qualitative study explored AMS work processes at two metropolitan hospitals within a single health district and uncovered significant differences across sites. Both hospitals conducted the same AMS tasks, however had different approaches to post-prescription review, interdepartmental AMS meetings, and tools used to support AMS, particularly Live AMS. Participants from both hospitals emphasised the importance of good working relationships between the AMS team and hospital prescribers, but also identified challenges associated with providing advice to prescribers. Organisational and cultural differences were found, with staff at Hospital B reporting they received inadequate support from hospital executive and senior staff, had increased prescriber autonomy, and required more resources. Interestingly, no themes were identified that related to the people element of SEIPS, which captures a person’s internal characteristics. Instead, results revealed that AMS work processes were dictated by factors external to people, including the interactions between people and other elements of the work system.
The approach taken for post-prescription review was a fundamental difference between the hospitals and impacted other elements of the AMS work system. Hospital A focused on all antimicrobials for their post-prescription reviews, and Hospital B focused predominantly on restricted antimicrobials. The availability of resources likely contributed to this difference, with participants at Hospital B reporting a lack of resources, and a shortage of ward pharmacists which impacted the AMS pharmacist’s time. Live AMS use may have also contributed to this difference, with participants at Hospital A observed using the tool comprehensively for post-prescription review, staff confirming that Live AMS was a key enabler of post-prescription review. As the current study did not evaluate patient outcomes, we are unable to determine the impact of each hospital’s approach to post-prescription review, however, participants held the view that increased review of antimicrobials would improve the scope of AMS and consequently patient outcomes. This is supported by previous research which has shown clinical pharmacist staffing levels, and inclusion of pharmacists in medical rounds, are associated with improved patient mortality.15 Further, participants at Hospital B identified issues and risks with unrestricted antimicrobials, including inappropriate continuation of prescriptions, and the need for a clinical decision support system to support antimicrobial prescribing, highlighting the value of having oversight of restricted and unrestricted antimicrobials.
Cultural differences emerged between the two hospitals, particularly within the AMS teams and their interactions with other departments. The notion of organisational culture includes the beliefs, values, norms, and behavioural patterns that define social relationships and are evident in an organisation’s structures and operational procedures.16,17 Hospital A’s AMS team felt well-supported, with involvement from all ID doctors and the AMS pharmacist working closely with the team. Hospital B participants lacked senior staff support, governance, and resources needed to enhance program outcomes. High levels of prescriber autonomy in Hospital B were also perceived to negatively impact the effectiveness of the AMS team. This prescribing culture could be attributed to the differences in interdepartmental meetings, with an increased number of meetings observed to occur at Hospital A, particularly with surgical specialties. The focus on surgical specialties is especially important due to their work processes in the hospital. A London teaching hospital study found distinct antimicrobial decision-making approaches between medical and surgical specialties.18 Surgical teams, dealing with an imbalanced skill mix on the ward and delayed senior input, frequently deferred de-escalation choices. Limited time and communication opportunities further hindered multidisciplinary teamwork.18 Therefore having dedicated time for AMS in surgical specialties is especially critical.
The interdepartmental AMS meetings were also reported to be a source of education and helped build rapport between departments. Interestingly, our results indicated differences in perceptions of interdepartmental relationships, with the hospital that had increased interdepartmental AMS meetings reporting they had good relationships with other departments. This is most likely attributable to the increased in-person interactions, as described in previous work exploring AMS team perceptions and activities occurring remotely in the context of COVID-19 restrictions.19 The effectiveness of in-person collaborative feedback in improving outcomes for AMS programs is well established in the literature.20-22.
Overall, the Live AMS dashboard was used more at Hospital A than B. Our results identified various factors which could account for this, such as advocacy from senior consultants, and the involvement of these consultants in the development of the system. Further, the lack of interoperability of the ICU eMM system at Hospital B meant that Live AMS could not be utilised for this department. Participants at Hospital B perceived Live AMS to not be useful for their tasks and were less aware of the features of the system. Interestingly, AMS work appeared to be completed successfully in the absence of Live AMS, perhaps indicating that the system was in fact not essential for all tasks. It is well recognised that AMS programs need to tailor strategies to their institution23, therefore it follows that the digital interventions used to facilitate these AMS programs should also be tailored. A prior understanding of workflows and user needs during implementation can also determine the expected usefulness and utilisation of a digital tool.24

Limitations

A strength of this study was its in-depth exploration of the work of AMS teams. However, the sample size was limited by the size of AMS teams, and the findings could be influenced by the characteristics and context specific to each team, potentially limiting the applicability of the results to other settings. Furthermore, the reliance on observations and interviews might introduce a degree of subjectivity and potential bias, which we attempted to overcome by having a second researcher independently develop the coding framework. As data collection was conducted during COVID-19 and conducted during a set period, the results may not capture the full extent of the AMS teams’ work processes over time (i.e. post-COVID).

Conclusions

In examining the work processes of AMS teams across two similar digital hospitals we found variations in the way tasks were performed and the contexts in which work was done. Organisational and cultural factors impacted the work processes of AMS with endorsement from executives, allocated resources, senior staff support, and frequent interdepartmental meetings reported and observed to enable effective AMS teams. These factors also contributed to improving AMS engagement, and the relationships between departments, which is vital to successful AMS. Further, digital system interoperability and the presence of user champions can promote user acceptance and utilisation of digital tools that may improve AMS outcomes.