Search results and included studies

We retrieved a total of 9186 citations from electronic databases and 100 from Google Scholar (Fig. 1). After removing duplicates, we screened titles and abstracts of 5522 publications, of which 441 went through full-text review and 52 were included in the final analysis. The majority were excluded for not explicitly reporting implementation barriers and facilitators.

Fig. 1
figure 1

Study characteristics and quality appraisal

Table 3 summarizes the characteristics of the 52 articles included in the review. Countries were classified based on the World Bank’s 2019 country classification [35], and we found that most studies were conducted in Sub-Saharan Africa (N=17, 33%) and East Asia and Pacific (N=16, 31%). About 75% of the studies were implemented in tertiary hospitals, followed by primary care settings (N=13, 25%), while community-based interventions were very few (N=4, 8%). Nearly half of the studies described interventions for improving the prescribing practice of physicians only (N=25, 48%), whereas interventions targeting users (N= 2, 4%) or both providers and users (N=3, 6%) were rarely reported. Additionally, 42% of the studies (N=22) described interventions that involved multiple providers, such as physicians, surgeons, pharmacists, and nurses. In terms of intervention types, most studies employed enablement (e.g., audit and feedback; N=35, 67%), restriction (e.g., developing and enforcement of guidelines; N=31, 60%), and education (e.g., information sessions; N=27, 52%) interventions, with only few studies used persuasion (N=4, 8%) and modeling (N=1, 2%) interventions. Additionally, in more than half of the studies, interventions included two or three components. The most commonly applied study designs were pre-post (N=17, 55%) and cross-sectional (N=15, 29%), while very few studies used more robust designs such as randomized control trials (N=1, 2%) or a quasi-experimental design (N=1, 2%). About 13% (N=7) of the studies used a mixed-methods approach. Detailed information of each study is summarized in Additional file 3.

Table 3 Characteristics of studies included in the review

Among quantitative studies that employed a non-randomized control trial design, most did not report complete outcome data or adjust for confounders in the analysis. Most studies that applied descriptive cross-sectional methods used sampling strategies inappropriate to address the research question. The nonresponse bias was also high across these studies. While most qualitative studies had relatively fewer biases, only one study had comparatively lower quality than the others. In many mixed-methods studies, the divergences and inconsistencies between quantitative and qualitative results were not adequately addressed. Three studies were not assessed using the mixed-methods appraisal tool, since they were narrative descriptions of the implementation process of an intervention or a conceptual model. Additional file 4 provides a breakdown of the quality assessment of each study.

Barriers and facilitators

We grouped the barriers and facilitators reported in the studies into the five domains and 39 constructs of the CFIR framework and tabulated the frequency with which constructs were addressed (Table 4). We only presented constructs that were addressed in at least five studies to ensure they are common challenges or facilitators shared across LMIC settings. Specific codes for barriers and facilitators based on the CFIR framework of each study are presented in Additional file 5.

Table 4 Overview of CFIR constructs that were addressed in studies as barriers or facilitators

Domain 1: intervention characteristics

Intervention source, which refers to the perception of key stakeholders about whether the intervention is externally or internally developed, was reported in six studies as facilitators [36,37,38,39,40,41]. In these studies, local stakeholders were involved in the development, implementation, or evaluation of interventions, which ensured their ownership and buy-in.

A total of 17 studies addressed evidence strength and quality as a factor for implementing behavior change interventions to promote rational antibiotic use. One shared challenge was the lack of reliable data on AMR patterns in LMIC settings, which is crucial to the development of localized recommendations and guidelines for an antibiotic prescription [39, 42,43,44,45,46,47,48,49]. Eight studies reported that stakeholders and target populations were more receptive to recommendations and guidelines that were developed by authoritative and credible sources or based on localized and reliable evidence [41, 43, 50,51,52,53,54,55].

Seven studies commented on the adaptability of an intervention that refers to the degree to which interventions were tailored to the local needs [37,38,39,40, 52, 54, 56]. Specifically, interventions that were designed to be embedded into routine practice, adapted to the local capacities and priorities, and delivered jointly by local and international teams were more likely to be accepted by target populations.

The design quality and packaging of interventions were reported in nine studies as a facilitator, as using innovative approaches or user-friendly tools to deliver interventions improved the intervention uptake [38, 49, 51, 53, 57,58,59,60,61]. For example, concise guidelines that clearly outlined the appropriate antibiotics for common infections and interactive training sessions were reported to be more receptive by prescribers. Furthermore, six studies reported that poorly designed interventions (such as lack of details in guidelines for antibiotic prescription and insufficient implementation time) or using ineffective approaches to deliver them (such as the development of guidelines without dissemination strategies) were barriers to achieving intended behavior change outcomes [36, 46, 50, 62,63,64].

Domain 2: outer setting

Eight studies reported patient needs and resources as a barrier to changing prescribing practice, as participants were often pressured by patients to prescribe antibiotics, despite the ongoing interventions [39, 48, 49, 51, 60, 62, 65, 66]. External policy and incentives were addressed in a total of 22 studies. Specifically, 12 studies found that it was difficult to promote appropriate behavior among prescribers and users in countries without national policies or guidelines for antibiotic use or those with weak enforcement of existing regulations [39, 40, 43, 48, 52, 63, 67,68,69,70,71,72]. In contrast, interventions that were developed in line with national AMR initiatives and the availability of national policies and guidelines facilitated the implementation of interventions to improve rational antibiotic use [38, 40, 46, 48, 52, 68, 70, 73,74,75].

Domaine 3: inner setting

A total of 30 studies reported barriers and facilitators in the structural characteristics construct, which refers to the organizational environment where interventions are conducted. Infrastructure constraints included insufficient laboratory capacity to provide data on AMR patterns and diagnostic results timely, lack of data management technology for auditing antimicrobial stewardship activities, lack of in-hospital pharmacies, and inadequate supply of high-quality and effective antibiotics [37, 39,40,41, 43, 44, 46,47,48, 52, 66,67,68,69,70, 73, 74, 76,77,78]. Furthermore, several studies reported that health facilities in LMICs did not have an established governance structure to lead antimicrobial stewardship programs [46, 48, 63, 67], and that high turnover of medical staff and bureaucracy within hospitals prevented the successful implementation of antimicrobial stewardship programs [40, 68]. On the contrary, improving the infrastructure of facilities where interventions were conducted, such as establishing microbiology laboratories and enhancing the supply of antibiotics, facilitated prescribers to make evidence-based decisions on treating patients with appropriate antibiotics [36, 43, 48, 50, 56, 60, 62, 68].

Networks and communications were reported in eight studies as a facilitator [37, 43, 47, 50, 56, 62, 79, 80]. Successful interventions were often developed and implemented by an experienced and well-coordinated team of local and international stakeholders. Various measures were taken to ensure good communication among implementers, participants, and higher-level stakeholders. For example, stakeholders held regular meetings that involved providing project updates and collaborative decision-making to facilitate project planning, implementation, monitoring, and evaluation, as well as keeping track of the achievement of set goals. Researchers also established platforms for efficient communication between medical staff, thus enabling optimal treatment of patients with antibiotics [43].

The norms and values of an organization, which are represented by the culture construct in the framework, were addressed in seven studies [41, 50, 58, 62, 68, 77, 81]. For example, interventions that were developed in the Western context and based on principles of democracy and teamwork were difficult to implement in organizations with a hierarchy culture [50]. In several studies that evaluated the outcomes of antimicrobial stewardship programs, a rigid hierarchy within institutions and disconnection between physicians and other medical staff prevented prescribers from accepting suggestions for antibiotic choices from junior staff or pharmacists who were responsible for auditing prescribing behaviors [41, 50, 68]. Additionally, tension and distrust between physicians and patients were reported as a barrier to educating patients about rational antibiotic use during medical consultation [81].

Leadership engagement refers to the degree to which leaders are committed to, involved in, and held accountable for the implementation of interventions. In 21 studies, the involvement and support of higher-level stakeholders (such as officials from the Ministry of Health, health authorities, experts, leaders in health facilities) and administrative staff were reported as a common facilitator for the successful implementation of behavior change interventions [37, 39, 40, 44, 48,49,50, 52, 54, 56, 58, 59, 62, 65, 70, 78,79,80,81,82,83]. The lack thereof had been documented to impede intervention implementation in seven studies [43, 46, 60, 63, 69, 71, 84].

In 23 studies, the common challenges faced by implementers and researchers were related to resources, including insufficient financial support for sustaining antimicrobial stewardship activities, shortage of human resources (such as microbiologists, pharmacists, and infectious disease specialists), and lack of technological support to facilitate the administration of interventions [38, 42,43,44,45,46, 48,49,50, 58, 62, 63, 67, 69,70,71, 73, 75,76,77, 79, 85, 86]. It was also commonly reported that prescribers participating in ASPs were often too busy to perform intervention activities. On the contrary, the availability of information technology for managing auditing data and patients’ records had reduced the workload of participants and improved the efficiency of managing antimicrobial stewardship programs. Researchers and implementers also leveraged locally available but often untapped resources to ensure that interventions could be sustained [44, 47, 50, 62, 73].

Access to knowledge and information about interventions was recognized as a facilitator in nine studies [41, 43, 51, 53, 55, 61, 65, 80, 83]. Implementers and researchers employed training, education sessions, and other publicity strategies to help participants access intervention information and familiarize with intervention activities and content.

Domain 4: characteristics of individuals

Eight studies reported barriers in the “knowledge and beliefs about the intervention” construct. In several studies, the target populations were not aware of the ongoing interventions. Sometimes, participants were concerned about the effectiveness of the intervention or unfamiliar with the intervention content [39, 63, 66, 68, 69, 73, 75, 76]. On the other hand, seven studies documented that participants were more receptive to behavior change interventions when they acknowledge these interventions to be important and useful in improving their awareness of AMR [41, 47, 49, 56, 68, 79, 81].

The individual stage of change was addressed in 11 studies as a barrier to changing prescribing practice [40, 41, 49, 59, 60, 63, 66, 68, 69, 71, 75]. Prescribers sometimes were reluctant or even resistant to change their habitual practice, because they were skeptical about the effectiveness of the interventions. In some cases, prescribers had already established perceptions around “best practices” for treatment, which was difficult to change.

Additionally, six studies reported that lack of motivation prevented behavior change of prescribers, as some were concerned about complaints from patients or reduction in salary if they refused to prescribe antibiotics [39, 48, 62, 78, 84, 86].

Domain 5: process

The engaging construct was addressed in 13 studies, which acknowledged that effective implementation of behavior change interventions to promote rational antibiotic use required the involvement and collaboration of a multidisciplinary team of medical staff, including physicians, clinicians, nurses, and pharmacists [44, 48, 54, 56, 57, 66, 70, 72, 73, 75, 79, 82, 87]. In eight studies, appointing a dedicated focal person, usually a pharmacist or an infectious disease specialist, for coordinating antimicrobial stewardship activities (such as ward rounds, auditing antibiotic prescribing behavior, and supporting intervention management) was a facilitator [36, 47, 52, 55, 61, 70, 74, 83].

Five studies acknowledged that regular monitoring and evaluation of interventions using robust methods helped implementers to identify gaps and areas for improvement [52, 56, 76, 83, 87]. However, this was difficult to achieve in settings where routine data for monitoring and evaluation was not available. Researchers and implementers were also concerned about the validity of outcome indicators being used for assessing the effectiveness of behavior change interventions [42, 44, 69, 79, 85].

Construct relationships

We found that several factors influencing the implementation of the behavior change interventions were interconnected as shown in Fig. 2. For example, insufficient laboratory capacity to provide data on AMR patterns (structural characteristics) hindered the development of context-specific guidelines on antibiotic use, which was commonly reported as a challenge in ASPs (evidence strength and quality) [39, 42, 43]. The lack of data management systems for auditing prescribing behaviors and antibiotic use (structural characteristics) led to insufficient capacity for monitoring and evaluating behavior change interventions (reflecting and evaluating) [42, 69, 79]. Another example is that engaging local stakeholders who were responsible for policymaking or program decisions (leadership engagement) facilitated the development of interventions to be incorporated into routine practice (adaptability), aligned with local priorities, and better fitted into the organizational culture (culture) [37,38,39,40]. It also helped to build ownership and buy-in from stakeholders (intervention source) and facilitate policy scale-up [36,37,38,39, 41, 88]. The engagement with stakeholders, as well as between intervention participants, was usually facilitated by good communication (networks and communications) [37, 43, 47, 50, 81]. Additionally, studies reported that patients sometimes pressured physicians to prescribe antibiotics (patients’ needs and resources), which might contribute to low motivation for changing prescribing behavior among prescribers (other personal attributes) [60, 62, 65, 81].

Fig. 2
figure 2

A map of the relationships between the constructs that were addressed in studies as barriers or facilitators

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit The Creative Commons Public Domain Dedication waiver ( applies to the data made available in this article, unless otherwise stated in a credit line to the data.


This article is autogenerated using RSS feeds and has not been created or edited by OA JF.

Click here for Source link (