Antimicrobial Stewardship at a Tertiary Center in Portugal: Insights from Prescribers

1. Introduction

Antimicrobial resistance (AMR) is a major public health threat worldwide. Various national and international entities have launched initiatives to prevent its emergence and to promote the rational use of antibiotics[1,2]. This concern was most recently reaffirmed in the 2024 United Nations High-Level Meeting on AMR[3], where Heads of State recognized AMR as one of the most pressing global health threats and emphasized the need to strengthen multisectoral responses through antimicrobial stewardship (AMS) and other coordinated efforts.

In Portugal, the National Program for the Prevention and Control of Infections and Antimicrobial Resistance (PPCIRA) was established in 2013[4] and is considered a national health priority. PPCIRA operates at central, regional (UR-PPCIRA), and local (UL-PPCIRA) levels. As per national legislation, local units must implement an Antibiotic Stewardship Program (ASP) to optimize antimicrobial therapy, prevent unnecessary or prolonged prescriptions, and reduce ecological impact.

While the structural design and objectives of ASPs are well established, their influence on clinical decision-making is less well understood. A recent survey conducted at our institution highlighted significant variability in prescribing practices across departments, shaped by local cultures, hierarchies, and individual attitudes toward stewardship[5]. These findings suggest that ASPs’ impact depends not only on protocols but also on contextual and behavioral factors.

At our tertiary center, the UL-PPCIRA, which was created in 2014, leads the implementation of a multidisciplinary ASP. Its key activities include the dissemination of local infection treatment protocols, generation of antimicrobial consumption and resistance reports, post-prescription validation of restricted antibiotics, automatic 7-day stop orders, ad hoc clinical consultations, and weekly multidisciplinary meetings as part of a structured support model for clinical departments. However, despite these efforts, the extent to which specific stewardship initiatives influence prescribing behavior remains largely unmeasured. Understanding how prescribers perceive and interact with ASPs is necessary for identifying areas for improvement. Moreover, such an assessment may provide a useful model for other institutions aiming to align ASP activities with clinical realities.

This study aims to assess physicians’ perceptions of the ASP through a structured questionnaire, with the ultimate goal of promoting more rational and evidence-based antimicrobial use. Additionally, beyond its local relevance, this study also aims to illustrate a feasible model for assessing prescribers’ perceptions of stewardship programs.

2. Materials and Methods

This was a cross-sectional study conducted between March and December 2024 at Hospital de São João, a tertiary university hospital in Northern Portugal with 1200 beds. The institution is highly differentiated, with active bone marrow, kidney, and heart transplant programs, a dedicated hematology-oncology service, and an advanced intensive care unit (ICU) that includes extracorporeal membrane oxygenation (ECMO) support.

The study population consisted of all the physicians involved in antibiotic prescribing decisions at the institution. A convenience sampling method was employed: all eligible physicians were invited to participate via institutional email. Participation was encouraged through monthly reminder emails sent to all eligible physicians over the course of the study period.

Data were collected using an anonymized online questionnaire developed by the research team and adapted from a French antimicrobial stewardship study[6]. The questionnaire, administered through Microsoft Forms, was designed to capture participants’ experiences with, opinions on, and perceived usefulness of the various interventions of the hospital’s ASP, as well as to identify potential barriers and facilitators to adherence. It comprised 39 items grouped into six domains: (1) knowledge of antibiotics; (2) awareness of local antibiotic prescribing protocols; (3) use of antimicrobial consumption reports; (4) awareness of the restricted antibiotics program; (5) engagement with unscheduled consultations on antimicrobial therapy; and (6) participation in scheduled weekly multidisciplinary case reviews. The tool was specifically developed for this study and has not been formally validated. Before dissemination, it was piloted with four resident physicians from different specialties at the same institution (not included in the final sample) to assess clarity, relevance, and comprehensiveness. These participants were selected for convenience, based on availability. Feedback from the pilot informed refinements to wording and structure. Additionally, members of the hospital’s ASP reviewed the final version for content appropriateness. The complete questionnaire can be found in the supplementary materials.

Descriptive statistics were used to summarize participant responses. Missing data were handled using pairwise deletion, excluding only the specific data points that were missing for each analysis. Data analysis was conducted using Microsoft Excel.

This study received prior approval from the hospital’s ethics committee. Informed consent was obtained from all participants after providing detailed information about the study objectives, procedures, potential risks, and benefits. Data protection measures included anonymization of all responses and secure storage of data, accessible only to members of the research team. Participation was voluntary, and respondents could withdraw at any time. The study was also conducted in compliance with the General Data Protection Regulation (GDPR).

3. Results

Out of 1,539 surveys sent through institutional e-mail, 154 valid responses were received, yielding an approximate 10% response rate. Most respondents were female (64,3%) and held a specialist positions (75,8%), mostly a medical specialization (64,9%) (Table 1). All predefined age groups were represented, with the highest proportion of respondents aged 25 to 35 years (30.6%).

3.1. Antibiotic Knowledge

Most respondents reported being able to manage simple infections only (59.6%). In terms of self-assessed knowledge, responses were predominantly classified as good or sufficient, with fewer respondents rating their knowledge as very good or insufficient (Table 2).

3.2. Protocols

Most respondents were aware of the existence of institutional protocols for antimicrobial prescribing (Table 3). However, the frequency of use of protocols was generally low. Among those who rarely or never used the protocols, the most commonly reported reasons included difficulty in locating the documents in the hospital’s shared folders (35.7%), reliance on pre-existing knowledge (18.8%), lack of time (13.6%), and a preference for national or international guidelines (9.7%). Other less frequent reasons included the perception that the protocols did not address the specific clinical question (9.1%), and concerns regarding whether the documents were up to date (5.8%).

3.3. Restrictive Measures

Most respondents were not aware of the existence of post-prescription validation of restricted antibiotics (Table 4). Additionally, awareness of which antibiotics were restricted was low. As for the impact of these restrictive measures on clinical practice, most respondents (69.5%) agreed or strongly agreed that the measures influenced their prescribing practices, while 17.3% were neutral or disagreed.

3.4. Unscheduled Consultations

Situations prompting unscheduled consultation with the antimicrobial stewardship team varied, with the most common reasons being managing unfamiliar infections and seeking expert opinion (Table 5). A significant number of respondents also reported requesting guidance when first-line antibiotics were not viable or when therapy needed to be adapted based on microbiological results.

Regarding the frequency of unscheduled contact with the ASP team, 14.9% of respondents reported never requesting support, 51.3% rarely, 22.1% monthly, 9.7% weekly, and 1.9% more than once per week.

3.5. Scheduled Weekly Multidisciplinary Meetings

Only 25.3% of respondents reported having scheduled case-based discussions on antimicrobial prescription (Table 6). Of those with scheduled meetings, most considered their duration adequate, with only two respondents suggesting the meetings were too brief. Similarly, the frequency of the meetings was deemed appropriate by most participants. Regarding the usefulness of these consultations, the vast majority (37 out of 39) found them beneficial for their clinical practice.

3.6. Overview of Interventions

When asked about the most useful antimicrobial stewardship interventions, respondents most frequently selected the availability of protocols, unscheduled support for specialties, and regular case discussions. Regarding preferred resources for reinforcement or implementation, the most commonly chosen were the inclusion of prescription recommendations in susceptibility reports, direct contact when antibiotic therapy could be improved, and structured training sessions on antimicrobial prescription (Table 7).

3.7. Feedback of Data on Antimicrobial Use or Antimicrobial Resistance

A total of 129 respondents (83.8%) expressed interest in receiving regular updates on antimicrobial consumption within their unit, while 10 (6.5%) were not interested, and 15 (9.7%) reported no opinion. Interest in receiving periodic information on local antimicrobial resistance patterns was reported by 142 respondents (92.2%), whereas 3 (2.0%) were not interested, and 9 (5.8%) had no defined opinion.

4. Discussion

This study explored how physicians perceive and engage with AMS activities at a tertiary hospital. Despite a modest response rate, the sample included a wide range of specialties and seniority levels, offering insight into AMS integration across the institution.

Four key findings emerged: (1) institutional protocols are underused, mainly due to accessibility issues; (2) regular weekly multidisciplinary case discussions are well received and could be expanded; (3) awareness of the hospital’s restrictive antibiotic policy is limited; and (4) there is high demand for resistance and consumption data among prescribers. Below, we discuss each of these findings and possible actions.

Most respondents knew protocols existed but used them infrequently. Barriers included poor accessibility, time constraints, and preference for national or international guidelines. This is consistent with findings from the French national survey our study was based on[6], where 76.7% of prescribers relied on guidelines issued by academic societies compared to 61.5% who used local hospital protocols. Similarly, a study among general practitioners in Germany[7] reported that only 39% frequently used antibiotic therapy guidelines. The survey results indicate that institutional protocols must be made more accessible and easier to navigate. Access to antimicrobial prescription protocols could be improved by reorganizing the hospital’s internal system to create a centralized and clearly labelled section that facilitates consultation, highlighting recently updated content. In addition to improving access, protocols may need to be updated and structured in ways that align more closely with clinicians’ decision-making and using robust and transparent methodologies.

Scheduled case discussions were viewed positively by respondents who had them within their departments, with the vast majority indicating that the frequency, duration, and clinical usefulness of these meetings were adequate. These results may support expanding scheduled discussions to more departments. While many AMS tools are automated or digital, the value placed on face-to-face, case-based discussions shows that clinicians still rely on personalized, real-time input for complex decisions regarding individual patients. Unscheduled consultations were less frequent and usually occurred when physicians faced difficult cases, which may reflect appropriate, case-driven use of the service rather than systematic integration into routine clinical practice.

Awareness of the hospital’s restrictive antibiotic policies was limited. Even among those who knew such measures were in place, many were unsure which antibiotics were affected. While most respondents agreed that these policies influence their prescribing decisions, the results suggest that communication around their scope and rationale could be improved. One possible approach could include sharing an updated and simplified list of restricted antibiotics with clinical departments, accompanied by brief explanations. Opportunities to reinforce this information might include departmental meetings or routine AMS interactions.

Interest in receiving regular feedback on antimicrobial resistance and consumption data was high. These findings support the potential value of systematic feedback strategies. However, previous research suggests that even when this information is available, applying it in clinical practice can be challenging[8,9]. In a recent study[9], only 26.9% of hospitalists reported using antibiogram data more than once a month, and decisions to prescribe antibiotics were often unaffected by susceptibility levels. Therefore, resistance data must be presented with interpretative guidance that is relevant to common clinical scenarios. Integration into existing prescribing tools may also facilitate its use.

The most valued ASP interventions included access to protocols, expert support, and therapeutic guidance embedded in microbiology reports. In general, prescribers tend to support educational and supportive interventions over restrictive measures6. Additionally, team dynamics and autonomy influence prescribing decisions, supporting the idea that ASP strategies should be adapted to local contexts5.

This study has some limitations. The response rate was low, which may introduce selection bias, particularly if those more engaged with stewardship were more likely to respond. The questionnaire was not formally validated, although it was adapted from a previously published tool and piloted locally. Responses were self-reported and may be subject to recall bias. Finally, as a single-center, cross-sectional study, the findings may not be generalizable and do not capture changes over time.

Despite these limitations, the study offers relevant insights for assessing ASP quality as perceived by the medical population and improving implementation. The inclusion of respondents from various specialties and training levels provides a broad perspective on prescribing behavior.

5. Conclusions

Four areas stood out for potential improvement: (1) making protocols easier to access and use; (2) maintaining and expanding scheduled multidisciplinary discussions; (3) improving communication around restricted antibiotics; and (4) delivering clear, actionable feedback on resistance and consumption data. In addition to its local relevance, this study offers a practical example of how AMS initiatives can be assessed from the perspective of prescribers. Understanding how they engage with different stewardship components can inform the design of more context-sensitive and effective interventions. By sharing our approach, we aim to contribute to ongoing efforts to integrate behavioral and organizational dimensions into the evaluation and refinement of ASPs in real-world settings. Future research could explore the impact of specific interventions longitudinally and examine behavioral drivers that influence prescribing across clinical areas.