Introduction

Why Understand the Relationship Between Safety Culture, Infection Prevention and Control, and Healthcare-Associated Infections

Given the major impact of healthcare-associated infections (HAI) on morbidity, mortality, and costs of care, there have been many efforts to prevent HAIs by improving infection prevention and control (IPC) practices in the last 20 years. These include establishing the U.S. Department of Health and Human Services (DHHS) National HAI Action Plan, required reporting of hospital HAIs to the Centers for Disease Control and Prevention (CDC) National Healthcare Safety Network (NHSN), state-based HAI prevention activities, Joint Commission National Patient Safety Goals specific to HAIs, public reporting of HAIs in Medicare data, and U.S. Agency for Healthcare Research and Quality (AHRQ) healthcare-associated infections program [1,2,3,4,5,6,7,8]. While substantial progress has been made, it remains difficult to ascertain the impact of individual initiatives on overall progress [9, 10].

Safety culture is one factor considered in recent years to have a major impact on patient outcomes including HAIs [11, 12]. The often-stated phrase “culture eats strategy for lunch (and breakfast and dinner)” exemplifies its presumed importance [13, 14]. Several initiatives are underway to improve culture related to patient and staff safety within healthcare settings [15,16,17]. Culture is also closely related to leadership [18,19,20].

What Is Safety Culture

Safety culture has been defined as the product of individual and group beliefs, values, attitudes, perceptions, competencies, and patterns of behavior that determine the organization’s commitment to quality and patient safety [21]. According to AHRQ, a culture of safety encompasses four key principles:

  • Acknowledgment of the high-risk nature of an organization’s activities and the determination to achieve consistently safe operations

  • A blame-free environment where individuals can report errors or near misses without fear of reprimand or punishment

  • Encouragement of collaboration across ranks and disciplines to seek solutions to patient safety problems

  • Organizational commitment of resources to address safety concerns [22].

The terms safety culture, safety climate, and organizational culture are conceptually distinct but often used interchangeably. Organizational culture refers to the deeply embedded norms, values, beliefs, and assumptions shared by members of an organization [23]. These elements evolve over time and are difficult to change. Organizational climate, by contrast, refers to the shared perceptions at a given point in time regarding organizational practices such as decision-making, advancement opportunities, and so on. These are more amenable to change. Safety climate is a subset (or microclimate) of overall organizational climate that focuses on people’s perceptions about the extent to which the organization values safety (for workers, patients, and/or the environment) [12, 24,25,26]. In this article, we use the broader term safety culture to include studies addressing safety climate and organizational climate when appropriate.

Why Safety Culture Should Be Measured

Healthcare quality oversight bodies and related groups recommend organizations address safety culture. Organizations accredited by The Joint Commission are expected to ensure that leaders maintain a culture of safety and regularly evaluate the culture using valid and reliable tools [27]. The National Quality Forum recommended measurement of culture in its Safe Practices for Healthcare [28]. Similarly, The Leapfrog Group, which conducts hospital surveys related to quality and safety, incorporates measurement and use of safety culture data in its ratings of hospital performance [29]. AHRQ also recommends yearly measurement of safety culture as one of its ten patient safety tips for hospitals [30]. The 2020 Institute for Healthcare Improvement (IHI) National Patient Safety Action Plan recommends assessing safety culture at least every 2 years [31].

The mechanism by which safety culture can influence the IPC process and HAIs has been described in several conceptual models, many of which derive from Donabedian’s model of structure, process, and outcome [25, 32]. Figure 1 presents a simplified conceptual model for the relationship between safety culture, IPC processes, and HAI rates derived from the previous work. Essentially, the input of individual characteristics (both patients and staff) interacts with organizational structural factors (such as staffing, equipment) and culture of safety (impacted by leadership) which then influences the process of care (e.g., adherence to recommended IPC processes such as contact precautions and hand hygiene). These inputs in turn collectively affect the outcome of HAIs for both patients and staff.

Fig. 1
figure 1

Conceptual framework for safety culture and infection prevention and control-related (IPC) process and outcomes

Full size image

There is evidence from the Comprehensive Unit-based Safety Program (CUSP) initiative that driving large-scale improvement in IPC practices and HAI rates requires multi-faceted strategies that address clinician knowledge attitudes and behaviors as well as organizational factors. The CUSP implementation framework includes engagement of frontline clinicians and institutional senior leadership; education in the science of safety; assessment of safety culture and infection control policies and procedures; regional consortia of stakeholders; expert coaching, consultation, and technical assistance; peer support; educational materials, tools, and webinars; data collection and feedback for performance monitoring; and patient and family engagement. Originally demonstrated in the Michigan Keystone Project Collaborative ICU initiative to reduce central line–associated bloodstream infections (CLABSI), AHRQ has supported the implementation and evaluation of the CUSP model nationally and its application to other types of HAIs (e.g., catheter-associated urinary tract infections (CAUTI) and ventilator-associated pneumonia (VAP)) and settings (nursing homes, dialysis) as well as other outcomes [33, 34]. Safety culture remains a key component of the CUSP initiatives.

There are additional reasons to understand the relationship between safety culture, IPC process, and HAIs. If a strong relationship exists, one can focus interventions on improving safety culture. In theory, an intervention that improves safety culture could also improve non-IPC-related outcomes (“lift all boats”). For example, a recent systematic review of 62 high-quality studies by Braithwaite et al. found that positive organizational and workplace cultures were associated with a variety of patient outcomes such as reduced rates of mortality, falls, and HAIs and increased patient satisfaction [35]. Though a discussion of strategies to improve safety culture is beyond the scope of this article, there is little research on the modifiability of culture beyond knowing it to be a slow process and difficult to accomplish [36,37,38,39].

Another reason to understand the relationship between safety culture, IPC, and HAIs is to establish the value (or lack thereof) of safety culture measurement and analysis activities. Collecting and analyzing safety culture data can require significant effort, burden, and expense [40, 41]. The evidence base is needed to help justify these efforts.

Research on Safety Culture and Outcomes

Previous systematic reviews have examined the relationship between safety culture and the IPC process and/or HAIs. These reviews have been generally supportive of the relationship between these factors but are based on limited evidence. One systematic review examined the relationship between patient safety climate and adherence to standard precautions. After excluding intervention studies, Hessels and Larson found only 7 studies eligible for inclusion, all of which were cross-sectional designs [42•]. A recent systematic review by van Buitjene included 20 studies and found 90% produced supportive evidence for the hypothesis of a link between organizational culture and HAI rates, despite the fact that 8 included studies did not actually measure safety culture [43••]. DeBono synthesized studies on the relationship between organizational culture and behavioral attitudes toward IPC [44••]. They found that effective IPC relies directly upon the successful interplay of multiple management systems strongly influenced by corporate culture, and that improving organizational culture appears to be a promising albeit challenging target for IPC improvement campaigns, although better-quality studies are urgently needed.

Fig. 2
figure 2

PRISMA flow diagram for literature review process

Full size image

Challenges to Linking Safety Culture, Infection Prevention and Control Activities, and Outcomes

There are several reasons why establishing a causal relationship between safety culture, IPC processes, and HAIs has been difficult in previous research. Challenges to measurement and analysis for each of these areas are described below.

Safety Culture

  • Studies that measure safety culture often use different instruments which vary in psychometric properties such as reliability, validity, and sensitivity to change [45].

  • Perceptions of safety culture are not necessarily related to actual practice. For example, a qualitative study by Szymczak eloquently describes how the situational complexity of interactions affects both the willingness and ability of staff to “speak up” to prevent an error [46].

  • Analyses of safety culture data may be too simplistic. For example, researchers debate whether to aggregate levels of agreement (consensus) within units and domains or to analyze data using dispersion models in which variability, the degree of disagreement among respondents, is more informative [47, 48].

Infection Prevention and Control Practices

  • Measuring adherence to IPC practices is not easy. There is no nationally standardized approach for hand hygiene measurement in the U.S.A., and interpretation of rates is often affected by threats to validity, such as the Hawthorne effect, and insufficient numbers of observations [49]. Measuring adherence to evidence-based care bundles using checklists, while more comprehensive, can be time consuming and variable in approach, timing, and numbers [50].

Healthcare-Associated Infections

  • Measurement of HAIs, though more standardized than IPC, is heavily dependent on time frames and surveillance processes. Some HAIs are more reliable to collect than others [51]. The unit of measurement—ICU level, hospital-wide, single HAI or multiple—also varies.

  • Factors that directly influence development of HAIs often differ according to the HAI of interest and are greatly affected by patient risk characteristics such as co-morbidities [52].

  • Establishing attribution for the HAI is challenging. As described by Gardam, during a hospital stay, a patient can be exposed to thousands of different environmental and human contacts that may lead to an infection [37]. While only one exposure may lead to infection, it is nearly impossible to determine which one was responsible.

Directionality and Causation

  • Directionality for cause and effect regarding IPC improvement and safety culture remains to be established. For example, can improving IPC processes result in enhanced safety culture or does improvement in safety culture result in better IPC processes?

  • The ability to establish relationships and causality is, of course, affected by the purpose of the initiative as well as the study design [53]. Articles are often based on uncontrolled, pre-post intervention QI projects never intended to be considered research studies [54,55,56]. However, some consider QI initiatives to be quasi-experimental designs when using pre-post or time series approaches, particularly when there is a control group [57]. While the common goal includes improving care processes and outcomes, QI research is distinct in its focus on broader, generalizable research questions with hypotheses, multiple sites, and advanced statistical analysis [58].

Purpose of This Review

In previous reviews, the association with outcomes has been limited by the research design and strength of the evidence. Authors rarely distinguished between reports of QI initiatives and research studies or explored safety culture measurement issues. For this review, we examined authors’ findings and interpretation of the relationship between safety culture, IPC processes, and HAIs in both QI and research articles. We also sought to quantitatively characterize the variation in measurement of safety culture, the IPC process, and HAIs. Though not intended to be a comprehensive scoping or systematic literature review, we performed a new search of PubMed to identify recent articles in this topic area.

Methods

Data Sources and Search Strategy

Two literature searches were performed by medical librarians. A structured literature search of PubMed was performed in November 2019 for relevant articles across all healthcare settings, published in English language in the past 10 years that included safety culture and infection control process and/or infection control outcomes. To broaden the search of safety culture, we also included terms such as safety climate and organizational culture. Shortly thereafter, a second broad non-structured search was also performed. The unstructured search utilized free text of terms “infection prevention and safety culture” and the “similar articles” function. The first search yielded 329 and the second yielded 114 abstracts and titles. An additional five records were identified from references in other articles. A total of 448 abstracts and titles were retrieved.

Eligibility Criteria, Screening, and Characterization Process

After de-duplication by the librarians, two researchers screened the 448 titles and abstracts for eligibility. We included studies that measured infection prevention processes or HAIs and also measured safety culture or included an intervention or objective to improve safety culture. We excluded commentaries, editorials, and related articles that were not studies or QI initiatives, and all articles involving healthcare organizations from outside the U.S.A. After 448 abstracts were screened, 99 full text articles were assessed for eligibility for inclusion and a total of 55 articles were included (Fig. 2).

Study data were collected and managed using a standardized data abstraction form in Research Electronic Data Capture (REDCap) electronic data capture tools hosted at the University of Iowa [59, 60]. REDCap is a secure, web-based software platform designed to support data capture for research studies, providing (1) an intuitive interface for validated data capture; (2) audit trails for tracking data manipulation and export procedures; (3) automated export procedures for seamless data downloads to common statistical packages; and (4) procedures for data integration and interoperability with external sources.

Five researchers were sequentially assigned to abstract information from the 55 articles. Reviewers documented safety culture measurement and domains; conceptual model used, if any; IPC process(s) measured; HAIs measured; study-reported findings and relation to safety culture; and added comments. All articles were also classified by type as being either QI or research; research studies were then categorized by study design as qualitative, cross-sectional, or quasi-experimental.

Each full-text article was reviewed by one researcher. However, when the researcher had doubts regarding the inclusion/exclusion criteria of an article or results were not clear, the article was independently reviewed by a second reviewer. Eighteen articles were reviewed by at least two reviewers. Disagreements between reviewers were discussed and resolved by group consensus. Since the review focused exclusively on published articles, it was not considered to be human subject research.

What Does the Research Show

Settings and Populations

Of 55 articles, most (49, 89.1%) were based on initiatives in hospitals, 6 of which focused on pediatric populations and 3 on veterans. Other settings included nursing homes (n = 3), dialysis clinics (n = 2), and dental offices (n = 1). The majority, 36 (65.5%), involved multiple sites and 19 were single-site studies.

Purpose and Design

When stratified by type, slightly more than ½ (28, 50.9%) were primarily QI initiatives, four of which were large scale involving more than 100 sites [33, 61,62,63]. The remaining 27 were research articles with the following designs: 17 cross-sectional, 6 quasi-experimental, and 4 qualitative.

Thirty-eight articles (69.1%) included an intervention, 15 of which utilized the CUSP approach.

Safety Culture Measurement

Articles varied considerably in measuring safety culture with respect to instruments, mode of survey administration, timing relative to an intervention (e.g., baseline only or multiple measurements over time), unit of analysis (e.g., hospital-wide or ICU specific), and respondents (e.g., all staff vs only infection prevention leaders). Overall, 39 (70.9%) articles measured safety culture with a standardized tool such as the AHRQ Hospital Survey of Patient Safety Culture (HSOPSC) (n = 16) [47, 64,65,66,67,68,69,70,71,72,73,74,75,76,77,78]; Nursing Home Survey on Patient Safety Culture (NHSPSC) (n = 2) [79, 80]; the Safety Attitudes Questionnaire (SAQ) (n = 9) [62, 81,82,83,84,85,86,87]; the National Database of Nursing Quality Indicators (NDNQI) (n = 2) [74, 88]; or qualitative methods (n = 4). Not all articles mentioned which safety culture domains were assessed, while some only assessed one domain of safety culture (e.g., teamwork) [78, 84].

Of the 39 articles, 25 (64.1%) measured safety culture at a single point in time (e.g., baseline) while 14 (35.9%) measured safety culture multiple times over the study period.

Most measurements occurred at a hospital level, but some were performed at a unit level: ICUs or NICUs (n = 11), surgical units (n = 2), cardiac surgery unit (n = 1), medical wards (n = 1), medical unit for the elderly (n = 1), and hemato-oncology unit (n = 1). Of note, eight articles reported information from hundreds of ICUs [47, 61,62,63, 69, 82, 89, 90] and at least two articles reported data from a wide variety of both ICU and non-ICU units [63, 68].

Most respondents to the safety culture assessments were physicians, nurses, and other staff working in the participating units. However, some articles focused exclusively on hospital epidemiologists or infection preventionists (such as [89, 91,92,93]), nurses (such as [64, 76, 94]), or members of a dental hygienist professional association [95].

Research articles measured safety culture more frequently than QI articles. All the 27 research articles measured safety culture compared with less than ½ of QI articles (12 of 28).

Outcomes of Interest: IPC Processes of Care and HAIs

Tables 1, 2, and 3 describe which aspects of IPC process were evaluated, what HAIs were measured, and the author’s findings and conclusions on the relationship between safety culture and IPCs and/or HAIs.

Table 1 Studies on the relationship between safety culture and infection prevention and control processes
Full size table
Table 2 Studies on the relationship between safety culture and healthcare-associated infections
Full size table
Table 3 Studies that address the relationship between safety culture, infection prevention and control processes (IPC), and healthcare-associated infections (HAIs)
Full size table

Table 1 presents 13 articles that address the relationship between safety culture and IPC processes. The IPC processes measured varied and included 4 articles that measured compliance with policies and guidelines ([64, 89, 92, 95]), two that measured compliance with HAI prevention bundles (e.g., CLABSI, daily interruption of sedation, antibiotic prophylaxis, standard precautions) ([83, 96]), and 3 articles that focused on hand hygiene ([81, 97, 98]). One article measured aspects of infection control structure [91]. Other process-related measures included willingness to adopt CUSP interventions [82], attitudes and beliefs about a Klebsiella pneumoniae carbapenemase (KPC)-producing Enterobacteriaceae control program [99], and how hospital CLABSI prevention efforts encouraged employees’ efforts to speak up and prevent errors [100].

Table 2 presents 20 articles that address the relationship between safety culture and HAIs. There was also substantial heterogeneity in the type of HAIs measured. However, most of them (n = 15, 75%) measured CLABSIs, CAUTIs, VAPs, or SSIs either exclusively or as a combined outcome [47, 61, 66, 68,69,70, 77, 79, 80, 86, 94, 101,102,103,104]. Other HAIs in single articles were colon SSIs [67], hospital-acquired sepsis [65], and C. difficile [77], and 3 did not specify the type of HAI measured [71, 84, 85].

Table 3 presents 22 articles that measured both IPC processes and HAIs to address the relationship with safety culture. In this group, most articles (14, 90.9%) measured process compliance with HAI prevention bundles (e.g., CLABSI, VAP, MRSA screening adherence) [62, 63, 72, 73, 75, 87, 88, 90, 93, 105,106,107,108,109] or hand hygiene adherence [110,111,112,113] or adherence to policy and guidelines [76, 78, 114]. Regarding HAIs, most (n = 17, 81.8%) measured CLABSIs, CAUTIs, VAPs, or SSIs either exclusively or as a combined outcome [62, 63, 72, 73, 75, 76, 78, 87, 88, 90, 93, 105,106,107,108,109,110, 113]. Of note, C. difficile was only explored as an outcome by two [63, 112], and two articles focused on organism-specific HAIs (e.g., MRSA and VRE) [63, 111]. Two articles did not specify the type of HAI measured [74, 114].

Relationship Between Safety Culture and Improvements in IPC Process and Patient Outcomes

Across all 55 articles (regardless of whether they measured or analyzed safety culture), 51 (92.7%) concluded there was an association between safety culture and IPC process and/or HAIs. A summary of the conclusions about the relationship between safety culture, IPC, and HAI is presented in Table 4.

Table 4 Summary of articles that related safety culture to infection prevention and control (IPC) processes or healthcare-associated infections (HAIs)
Full size table

Of the 39 that measured safety culture, only 26 (67%) tried via statistical analysis to establish an association between safety culture and IPC process or HAIs. Statistical analyses varied by design.

Among the 28 QI projects, only 12 measured safety culture and 3 tried to establish a statistical association between safety culture and IPC process or HAIs. Nevertheless, almost all QI projects (27 out of 28, 96%) concluded there was a relationship between safety culture and outcomes. One of the 3 QI projects that measured safety culture and tried to establish a statistical association between safety culture and outcomes did not find a statistically significant relationship between culture and IPC process or HAIs [73].

By comparison, all 27 research articles measured safety culture. Of these, 23 (85.2%) tried to establish a statistical association between safety culture and IPC process or HAIs. Nineteen of the 23 (82.6%) identified a statistically significant association with IPC process, HAIs, or both. Among those that did not find an association were 2 of the largest quasi-experimental studies [68, 80] and 1 cross-sectional study [83]. Another smaller quasi-experimental study with an intervention exclusively focused on improving safety culture also did not find significant association [86].

Discussion

The goal of this review was to examine recent literature on the relationship between safety culture and IPC processes and outcomes in U.S. healthcare organizations. In so doing, we also sought to characterize the challenges to empirically establishing these relationships as well as limitations of current research.

The 55 articles reviewed varied considerably regarding purpose, design, safety culture measurement tools and respondent, and outcome (IPC process and/or HAI) measurement, although several were based on the CUSP model. This heterogeneity contributes to the previously described challenges to linking safety culture, IPC processes and outcomes.

Nevertheless, most of the QI and research articles that measured safety culture and statistically analyzed the associations concluded that safety culture was positively associated with IPC processes and negatively associated with HAIs. Thus, our findings are generally consistent with other reviews supporting the notion that an association exists [42•, 43••, 44••].

Interestingly, a greater proportion of QI articles than research articles concluded there was a relationship with safety culture. All but one of the QI projects concluded that safety culture improved over time, even though less than ½ of the QI projects actually measured safety culture.

Why is it that so many QI studies concluded there was a relationship even in the absence of measuring safety culture? Do the QI studies share a common assumption that safety culture must have improved? In the absence of empirical data, should we dismiss all the QI studies that did not measure culture and limit conclusions to the traditional research studies? Were this a systematic review that only included studies with rigorous designs, we would have to conclude there is limited evidence to support the relationship between safety culture and IPC and HAIs because of weak designs.

We propose an alternative explanation. Perhaps one should instead accept the experience reported by authors that culture did change, even if not measured in a quantifiable way. As described in the tables, authors often expressed that they saw or felt a culture change which was integral to the effectiveness of the improvement effort. In fact, it is quite possible that culture did improve in certain domains because of the increased attention of leadership, extra resources, incentives, and momentum to drive change. Awareness and attitudes about the preventability of HAIs likely changed through education, and a sense of teamwork likely improved through the shared experience of working toward a common, measurable goal. This notion is highly consistent with the multi-faceted CUSP approach to improvement initiatives.

On the other hand, it is important to note that some research studies with the strongest methodologies (large numbers of sites, multiple outcomes, and longitudinal design) did not find an association between safety culture, IPC process, and HAIs (e.g., [68, 80]). Meddings et al. suggest that improvements in HAIs occurred by means other than improving safety culture, such as a strong emphasis on standardizing technical components of care, and that changing safety culture may not be critical to improvement [68]. Alternatively, they also suggest that safety culture may have improved and been instrumental in improving outcomes but could have been inadequately measured. In response, others have argued that the conclusions drawn by Meddings were unrelated to the instrument and inadequately supported by the methods and analyses [115].

So, what are the practical implications of the findings? Should one continue measuring safety culture and trying to establish relationships with outcomes? What does this say about targeting interventions to improve safety culture? Some insight into this question comes from the review by De Bono [44••]. They state that strategies aimed at achieving long-term improvement in IPC performance need to be multi-modal in order to both reflect and address the multidimensionality of the structure and the dynamic of organizational culture. They state “the link between compliant behavior and culture may seem to be plausible and sometimes obvious. But when empirically measured, the relationship might be weaker than expected, probably due to a long chain of mitigating factors and confounders.”

De Bono and colleagues also argue that strategy and organizational culture are inextricably intertwined in that all attempts at organizational culture change should be viewed as strategic changes [44••]. The notion that strategy and culture are inextricably intertwined helps explain why so many healthcare organizations in this review experienced a culture change while implementing strategies for improvement. Regarding causal direction, it may well be that implementing effective QI intervention strategies influences culture as much or more than interventions to change culture affect IPC processes and HAIs.

Regardless of the inter-relationships between safety culture, process, and outcome, there remains value in assessing safety culture. This value includes the ability to better understand staff perceptions and attitudes, identify opportunities for improvement, find differences across units and healthcare worker roles, comply with recommendations and requirements of oversight groups, and effectively implement interventions that ultimately benefit staff and patients.

Research Gaps

We identified several gaps in need of future research. Foremost is the need for longitudinal, cluster randomized trials with the explicit purpose of assessing the impact of safety culture, while controlling for the effect of specific interventions. Greater rigor in design would overcome many of the weaknesses inherent in uncontrolled observational studies. More research is also needed in non-hospital settings. It would be interesting to know if the relationships would be similar, stronger, or weaker in settings that are less organizationally complex than hospital, such as nursing homes and specialty clinics.

Similarly, more studies that directly measure safety culture over time are needed to better understand the sensitivity to change of safety culture instruments. Greater attention to fine tuning analysis for specific subgroups is also needed. For example, do associations between safety culture and IPC process vary depending on the role of the person completing the safety culture assessment (frontline staff vs leadership), the domains within safety culture (e.g., teamwork, communication, leadership), or the HAI of interest? Also, to what extent do studies at the unit-level generalize to organizational level findings?

Limitations

Our review has several important limitations. It was not intended to be a systematic review, and we did not formally evaluate the quality of studies. Nor should it be considered a comprehensive scoping review because we did not strive to assess the breadth of literature on this topic. We did not include gray literature that was not peer reviewed.

We only searched one citation database (PubMed). Other databases such as CINAHL, PsycINFO, EMBASE, and Cochrane might have yielded more studies. We excluded studies done outside the U.S.A. in part because of differences in facility characteristics and payment systems across borders which could confound the relationships of interest. There is a substantial body of work on this topic in European, Asian, and other regions which might have led to different conclusions (e.g., [116, 117]). The choice of search terms and PubMed field types may have been too limited. We also may have missed studies due to publication bias in that studies lacking associations between safety culture and IPC process or HAIs may never have been submitted or published.

Finally, in our analysis, we did not consider other related factors such as leadership styles and issues specific to the measurement tools. Admittedly, lumping together tools that measure culture and climate further conflates the distinction in instruments and does nothing to address psychometric concerns related to validity, reliability, and domain-specific issues.

Conclusions

Safety culture appears to be related to IPC processes and HAI outcomes. The associations with the IPC process may be bi-directional in that positive safety culture contributes to the success of interventions and implementing interventions drives improvements in safety culture. In the absence of longitudinal controlled trials, it may not be possible to separate the effect of safety culture from that of implementing effective intervention strategies on HAIs. Nevertheless, there remains value in evaluating safety culture in order to understand staff perceptions, identify opportunities for improvement, and implement interventions that ultimately benefit staff and patients.