Systematic review: Nurses' safety attitudes and their impact on patient outcomes in acute‐care hospitals

Abstract Aims The aim of this review was to synthesize the best available evidence on the impact of nurses' safety attitudes on patient outcomes in acute‐care hospitals. Design Systematic review with a narrative synthesis of the available data. Data sources Data sources included MEDLINE, Cumulative Index of Nursing and Allied Health Literature, Scopus and Web of Science Core Collection. Studies published up to March 2021 were included. Review Methods This review was conducted using guidance from the Joanna Briggs Institute for Systematic Reviews and reported as per the Preferred Reporting Items for Systematic Review and Meta‐Analyses guidelines. Results A total of 3,452 studies were identified, and nine studies met the inclusion criteria. Nurses with positive safety attitudes reported fewer patient falls, medication errors, pressure injuries, healthcare‐associated infections, mortality, physical restraints, vascular access device reactions and higher patient satisfaction. Effective teamwork led to a reduction in adverse patient outcomes. Most included studies (N = 6) used variants of the Hospital Survey on Patient Safety Culture to assess nurses' safety attitudes. Patient outcomes data were collected from four sources: coded medical records data, incident management systems, nurse perceptions of adverse events and patient perceptions of safety. Conclusion A positive safety culture in nursing units and across hospitals resulted in fewer reported adverse patient outcomes. Nurse managers can improve nurses' safety attitudes by promoting a non‐punitive response to error reporting and promoting effective teamwork and good communication.

| 31 ALANAZI et AL. patient outcomes (Han et al., 2020). These attitudes are referred to as safety attitudes and relate to an individual's beliefs, perceptions, feelings or thinking towards safety practices, procedures and policies (Sexton et al., 2006). When individual nurses have positive safety attitudes, a strong safety culture develops .
A positive safety culture in healthcare involves reporting of errors, a non-punitive response to error, management support and commitment to investigate error, effective communication and organizational learning (Lee et al., 2019). Conversely, fear of reporting errors is an indicator of poor safety culture as the identification and investigation of errors are considered critical in improving patient outcomes by developing strategies aimed at preventing reoccurrence of similar incidents (World Health Organization, 2021). Therefore, nurses' safety attitudes may have a significant impact on patient outcomes and patient safety.

| Background
Hospital or unit safety culture consists of the collective safety attitudes of staff. A commonly used definition of safety culture is "the product of individual and group values, attitudes, perceptions, competencies, and patterns of behaviour that determine the commitment to, and the style and proficiency of, an organization's health and safety management" (Waterson, 2014, p. 68-69). The term "safety climate" is frequently used interchangeably with the term "safety culture" (Hogden et al., 2017). Individual safety attitudes are most frequently measured by Likert scale self-report questionnaires . A number of instruments are used to evaluate safety culture and/or safety attitudes in healthcare settings (Hogden et al., 2017). The most commonly used instruments are the Hospital Survey on Patient Safety Culture (HSOPSC) and the Safety Attitudes Questionnaire (SAQ; Ellis et al., 2020). Both the HSOPSC and the SAQ are valid and reliable and have been translated into multiple languages (Hogden et al., 2017;Okuyama et al., 2018). For the purpose of this review, the term safety attitudes was used to describe an individual nurse's safety attitudes and aggregated scores at the unit or hospital level will be referred to as safety culture.
Patient outcomes can be defined as the changes in, or maintenance of, the patient's health-related state and include measures such as adverse events and patient satisfaction with care . Patient outcomes data can be obtained from administrative datasets, which include incident management systems and discharge datasets, and from staff and/or patient surveys (Alghraiybah et al., 2021;Sim et al., 2018). Data obtained from administrative datasets are considered the "gold standard" because they use routinely collected data and can include large sample sizes (Sim et al., 2019). However, nurse perceptions of the frequency of adverse events are also widely used as nurses are reliable providers of information about patient outcomes and are aware of patient safety issues (Lake et al., 2016;. Previous research has identified that nurse staffing and the nursing work environment can impact on patient outcomes Stalpers et al., 2015). Nurses' safety attitudes are also thought to influence patient outcomes (Han et al., 2020).
The RN4CAST research programme in Europe (conducted in 243 hospitals in six countries) identified that one in three nurses (N = 13,077 nurses) were reported to have poor safety attitudes . Other studies have identified that nurses' poor safety attitudes are associated with negative patient outcomes (Han et al., 2020;. Given that nurses comprise the largest percentage of the hospital workforce (Sim et al., 2019), the impact of nurses' safety attitudes on patient outcomes requires further study.

| Aim
The aim of this review was to synthesize the best available evidence on the impact of nurses' safety attitudes on patient outcomes in acute-care hospital settings. The question guiding this review was: "what is the impact of nurses' safety attitudes on patient outcomes in acute care hospitals?"

| Design
This systematic review was conducted in accordance with guidance from the Joanna Briggs Institute (JBI) for Systematic Reviews  and reported as per the Preferred Reporting Items for Systematic Review and Meta-Analyses (PRISMA) guidelines (Page et al., 2021). The PICO elements (Population; Intervention; Comparator; Outcome) were used to formulate the review question (see Table 1) . A search of the PROSPERO database identified that no systematic reviews on this topic had been previously conducted or were currently in progress. The review protocol was registered a priori with PROSPERO in July 2020 (CRD42020159074).

Impact Statement
What does this paper contribute to the wider global clinical community?
• Nurses with positive safety attitudes report fewer adverse events; and units and hospitals with a positive safety culture have improved patient outcomes.
• Effective teamwork and communication among nurses lead to a reduction in the occurrence of adverse patient outcomes.
• Nurse managers play an important role in promoting a positive safety culture which in turn improves nurses' safety attitudes and ultimately patient outcomes.

| Search method
The search strategy consisted of two phases. The first involved searches in MEDLINE (EBSCO Host), Cumulative Index of Nursing and Allied Health Literature (CINAHL, EBSCO Host), Scopus and Web of Science Core Collection. The second phase examined the reference lists of all included studies to identify any additional studies relevant to the review. The following search terms were used in all databases: "safety attitude*" OR "safety culture" OR "safety climate" AND "nurs*" AND "patient outcome*". The search strategy for CINAHL has been documented in PROSPERO. The initial search was conducted with the assistance of a health librarian in 2020 and then updated in March 2021.

| Inclusion and exclusion criteria
Studies were included if they were published in English, collected data from nurses working in acute-care hospitals and explored the impact of nurses' safety attitudes on patient outcomes. Studies that examined nurses' safety attitudes as part of the interdisciplinary team were included if data from nurses could be extracted. Studies examining safety attitudes, safety culture or safety climate were included. In this review, acute-care hospitals were defined as general, non-specialist hospitals with surgical and medical facilities that provide active diagnosis, care and treatment of a broad range of acute conditions including trauma and injuries (Hirshon et al., 2013). No restrictions were applied to qualifications, education or experience of nurses in the included studies.

| Search outcomes
The search identified 3,452 studies. EndNote was used for bibliographic management (The EndNote Team, 2013). The screening process was conducted in sequential steps (see PRISMA flow diagram, Figure 1; Page et al., 2021). After removal of duplicates, a total of 2,866 studies underwent title and abstract screening by one author (FKA) and 2,814 were not relevant and were excluded. The remaining fifty-two full-text studies were then independently reviewed by all authors and nine were deemed eligible for inclusion. No disagreements arose over inclusion and exclusion criteria, and the nine studies were retained for quality appraisal.

| Quality appraisal
The quality of included studies was assessed using the JBI critical appraisal tool for analytical cross-sectional studies. The tool includes eight questions with response options of Yes, No, Unclear or Not applicable (Moola et al., 2020). All appraisal scores were converted to a percentage, and scores of yes for ≥70% of questions were considered low risk of bias; 50%-69% moderate risk of bias; and ≤49% high risk of bias (Melo et al., 2018). All three authors critically appraised each study independently and then met to compare their evaluations. There were no disagreements, and all nine studies were included in the review. Quality appraisal results are documented in Table 2, and the checklist is provided in Appendix S1.

| Data abstraction
Data were extracted from each study into a summary table (Table 2) by two authors (FKA and JS) independently. This included details on the citation (year) country, design and sample, risk of bias, safety culture measures, patient outcomes measures and key findings relating to safety attitudes and culture.

| Data analysis and synthesis
A meta-analysis was not feasible due to heterogeneity in how patient outcomes were defined and examined and variation in methodological approaches used to collect both safety attitudes and patient outcomes data. Therefore, data were synthesized narratively using the relevant items from the Synthesis Without Meta-analysis (SWiM) guidelines (Campbell et al., 2020). Data related to nurses' safety attitudes were compared and summarized using six domains.
Patient outcomes were grouped and narratively synthesized.

| Characteristics of included studies
Of the nine included studies, five were undertaken in the United States of America (USA; N = 5), with one each from Canada, Switzerland, South Korea and China. All studies used cross-sectional designs.

Population
Nurses in acute-care hospitals

| Safety attitudes of nurses
All included studies rated safety attitudes as positive. The highest overall mean score for safety attitudes was 3.94 (out of 5), SD= 0.44 (Brown & Wolosin, 2013) using the 42-item version of the Hospital Survey on Patient Safety Culture (HSOPSC). The lowest mean score was 55.0%, SD = 8.7%, using the seven-item version of the HSOPSC (Olds et al., 2017). A positive score on teamwork elements was identified in six studies (Ausserhofer et al., 2013;Brown & Wolosin, 2013;Han et al., 2020;Hessels et al., 2019;Taylor et al., 2012;Wang et al., 2014). One study reported that teamwork and collaboration between staff were sub-optimal (Olds et al., 2017). Stress recognition was examined in three studies and all reported positive scores (Han et al., 2020;Hessels et al., 2019;Wang et al., 2014). Most studies reported that management provides appropriate responses when errors are reported (Ausserhofer et al., 2013;Brown & Wolosin, 2013;Han et al., 2020;Hessels et al., 2019;Hofmann & Mark, 2006;Taylor et al., 2012;Wang et al., 2014). However, one study identified that although patient safety was perceived to be a top priority of management, nurses were scared of making mistakes and feared punishment from management which may lead to some errors not being reported (Olds et al., 2017).  -Mean level of unit safety culture was 5.11 (0.49), 7-point scale (higher scores indicate better safety culture).
-There was no significant relationship between safety culture and any patient outcomes. • Patient falls.

Brown and Wolosin
• Falls with injury (Moderate Injury or above).
-Mean level of unit safety culture was 3.94 (0.44), 5-point scale (higher scores indicate better safety culture).
-Stronger overall safety culture led to fewer pressure injuries r TA B L E 2 (Continued) Climate Scale (ZSCS) and Error Orientation Scale (EOS), were also used to measure nurses' safety attitudes. The SOS includes nine items that assess nurses' engagement in safety behaviours at the unit level (Ausserhofer et al., 2013). The SAQ used by Taylor et al. (2012) included 36 items and five subscales. One study used both the ZSCS (9 items) and the EOS (13 items) to measure nurses' perceptions about safety culture (Hofmann & Mark, 2006). All instruments measured safety attitudes using five-or seven-point Likert scales.
Results were presented as either an overall mean score, composite scores for each domain or converted to a percentage. Higher scores indicated a stronger safety culture or safety attitude in all studies.
The characteristics of the instruments used to assess nurses' safety attitudes are presented in Table 3.

| Patient outcomes data
The

| Patient falls
Seven studies examined patient falls and three reported significant associations between nurses' positive safety attitudes and a reduction in patient falls (Brown & Wolosin, 2013;Han et al., 2020;. Brown and Wolosin (2013) identified that nursing units with positive teamwork had lower numbers of patient falls recorded at the unit level (r = −0.32, p < .05). In addition, higher levels of management support for patient safety led to nurses accurately reporting the patient falls that occurred (r = +0.35, p < .03) (Brown & Wolosin, 2013). This indicates that when nurse managers create sup-  -Mean level of individual safety attitude was 3.46 (0.60), 5-point scale (higher scores indicate better safety culture).
-Negative relationship between stronger safety culture subscales and fewer patient outcomes (except patient falls).

| Pressure injuries
Of the five studies that examined pressure injuries, four stud-

| Mortality
Only one study examined nurses' safety culture and patient mortality. Olds et al. (2017) identified that positive safety culture was associated with a decreased risk of death (OR = 0.92, CI = 0.89-0.95).

| Physical restraints
The use of physical restraints was examined in two studies (Han et al., 2020;Wang et al., 2014) using nurse perceptions of adverse events. Nurse reports of physical restraints being used for more -Staff feel like their mistakes are held against them.

TA B L E 3 Characteristics of the instruments used to assess nurses' safety attitudes in the included studies
-Important patient care information is often lost during shift changes.
-Things "fall between the cracks" when transferring patients from one unit to another.
-Staff feel free to question the decisions or actions of those in authority.
-In this unit, we discuss ways to prevent errors from happening again.
-We are given feedback about changes put into place based on event reports.
-The actions of hospital/organization management show that patient safety is a top priority.  -We talk about mistakes and ways to learn from them.

5-point Likert scale
-We discuss our unique skills with each other so we know who on the unit has relevant specialized skills and knowledge.
-We discuss alternatives as to how to go about our normal work activities.
-When giving report to an oncoming nurse, we usually discuss what to look out for.
-When attempting to resolve a problem, we take advantage of the unique skills of our colleagues.
-We spend time identifying activities we do not want to go wrong.
-When errors happen, we discuss how we could have prevented them.
-When a patient crisis occurs, we rapidly pool our collective expertise to attempt to resolve it.

7-points Likert scale
Likert scale range from 1 = not at all, 2 = to a very limited extent, 3 = to a limited extent 4 = to a moderate extent 5 = to a considerable extent 6 = to a great extent 7 = to a very great extent Mean score above 3.5 indicates a positive safety culture. -Job duties (three items).

5-point Likert scale
Range from 1 = strongly disagree to 5 = strongly agree. Mean score above 2.5 indicates a positive safety culture.

TA B L E 3 (Continued)
than eight hours reduced when nurses worked in hospitals that had optimal organizational learning (OR = 0.40, p = .019) and effective communication (OR = 0.54, p = .010; Wang et al., 2014). Similarly, restraints use for more than eight hours decreased when patient safety was promoted by managers (OR = 0.39, CI = 0.19-0.79) and when there was increased collaboration between staff during patient handovers (OR = 2.02, CI = 1.02-3.98; Han et al., 2020).

| Vascular access device
The correlation between vascular access device complications and nurses' safety attitudes was examined in two studies (Hessels et al., 2019;Wang et al., 2014). Wang et al. (2014) found that a higher mean score for handovers and transitions in

| D ISCUSS I ON
The findings from this systematic review indicate that nurses' positive safety attitudes lead to improved outcomes for patients in acute-care hospital settings. Despite purposely searching for studies that examined patient outcomes, most included studies examined adverse events in isolation from other outcome measures. This focus on adverse events among included studies is also evident in the literature on nursing-sensitive patient outcomes (Sim et al., 2018) and may relate to the availability of data and the presence of valid metrics to quantify these events. Only four of the nine included studies explored patient satisfaction (Ausserhofer et al., 2013;Hofmann & Mark, 2006;Wang et al., 2014) and quality-of-care issues (Hessels et al., 2019) within this review. Other indicators such as length of stay, failure to rescue and readmissions may also be influenced by nurses' safety attitudes; however, the absence of data on these topics was surprising.
Nurses play a key role within healthcare teams, and the safety Consequently, financial penalties are increasingly being imposed on healthcare providers to acknowledge the gravity and preventability of pressure injuries (Nguyen et al., 2015). While a number of initiatives such as environmental modifications (Clay et al., 2018), Nurses' positive safety attitudes and reduction in mortality, use of physical restraints, vascular access device complications and higher patient satisfaction were also examined within the included studies. The findings on these patient outcomes are limited in that they were based on a small number of studies; however, some of the associations are consistent with previous literature on the nursing practice environment. A systematic review and meta-analysis reported that positive nursing work environment was associated with a lower probability of mortality and higher patient satisfaction (Lake et al., 2019). Another systematic review found an association between strong working conditions and a reduction in the rate of mortality; however, there was no association with patient satisfaction (Bae, 2011). Considering a few studies have examined the impact of nurses' safety attitudes on these patient outcomes, these outcomes require further investigation.
The findings from this systematic review are based on studies that were observational in nature, with many studies relying on selfreported data. In addition, nurses' safety attitudes were assessed using a variety of different tools and at different organizational levels. Five validated instruments were used in the nine included studies. These differences in how nurses' safety attitudes were measured led to inconsistencies, which limit our understanding of the association between nurses' safety attitudes and patient outcomes.
Furthermore, analysis at different organizational levels (individual nurse, nursing unit/ward and hospital) further limits our understanding of these associations. Two studies that analysed safety attitudes at the individual level, using the full version of HSOPSC and nurses' perception of adverse events, found statistically significant relationships (Han et al., 2020;Wang et al., 2014). Hospital-level analysis was used in one study using the seven-item version of HSOPSC and nurses' perception of adverse events  and another using administrative data (Olds et al., 2017) and both studies identified statistically significant relationships. In contrast, studies that used unit-level analysis reported limited associations between nurses' safety attitudes and patient outcomes. For example, patient falls were not significantly associated with the nursing unit safety culture in three studies (Ausserhofer et al., 2013;Hessels et al., 2019;Taylor et al., 2012). This contrasts with studies that evaluated patient falls at the individual (Han et al., 2020) and hospital level . Interestingly, individual safety attitudes and nurses' perceptions of adverse event data remained consistent when the data were aggregated to the hospital level.

| LI M ITATI O N S
Although a comprehensive search strategy was used, it is possible that not all relevant studies were included in this review. More than half of the studies were conducted in the USA; therefore, further international research is required. Wide variation in the number of instruments used, and the source of patient outcomes data was noted.
For instance, data related to the frequency of patient falls was collected using a cross-sectional survey in some studies (Ausserhofer et al., 2013;Han et al., 2020;Wang et al., 2014) and via administrative datasets in others (Brown & Wolosin, 2013;Hessels et al., 2019;Taylor et al., 2012). This heterogeneity impacts upon our understanding of the complex relationship between nurses' safety attitudes and patient outcomes. In addition, nurses' perceptions of adverse events may also be influenced by how the individual nurse's perceptions of outcomes were measured and analysed. Two included studies (Han et al., 2020;Wang et al., 2014) dichotomized participant responses into groups of "never happened" or "happened" (thereby comparing "never happened" with the grouping of the other six categories including "happened several times a year" to "happened daily") during data analysis. Other studies dichotomized the frequency of events so that events that never occurred or rarely occurred (once per month or a few times per year) and those that occurred occasionally or frequently were grouped together (a few times per month, once a week, a few times a week and every day) Lake et al., 2016). These different approaches led to variability in results between studies and made the synthesis of findings challenging. Furthermore, all of the included studies were cross-sectional in design, which makes it difficult to make causal inference between nurses' safety attitudes and patient outcomes.

| CON CLUS ION
Evidence from the included articles shows that acute-care hospitals and units with positive safety culture report fewer adverse patient outcomes. These findings suggest that nurses' positive safety attitudes have the potential to improve patient outcomes and prevention of harm to patients in acute-care hospital settings. Individual nurses must be empowered to develop and maintain positive safety attitudes as these attitudes contribute to the safety culture of the ward or unit where they work. The collective safety culture of a ward is influenced by the nurse managers approach to safety. However, future research to fully explore the relationship between nurses' safety attitudes and patient outcomes using stronger research designs is warranted.

| RELE VAN CE TO CLINI C AL PR AC TI CE
This systematic review highlights that nurses with positive safety attitudes substantially improve patient safety in acute-care settings.
Teamwork, collaboration and communication among nurses in the area of professional practice lead to a reduction in the occurrence of adverse patient outcomes. Nurse managers can improve patient outcomes by focusing on improving nurses' safety attitudes. Nurse managers play essential roles in building and promoting nurses' safety attitudes through having a non-punitive response to error, creation of trusting relationships and developing environments where individuals can learn from error.

ACK N OWLED G EM ENTS
The authors would like to thank Dr Celeste Rossetto, Learning Development at the University of Wollongong, for responding kindly to queries about the writing.

CO N FLI C T O F I NTE R E S T
No conflict of interest has been declared by the authors.

AUTH O R CO NTR I B UTI O N S
FKA involved in searching databases, collecting data and writing of the initial draft. FKA, JS and SL independently screened and apprised the full text included studies. JS and SL comprehensively revised the content of review after each draft. All authors approved the final version to be submitted for publication.

E TH I C A L A PPROVA L
This was a systematic review of previously published research, and thus, Research Ethics Committee approval was not required.

DATA AVA I L A B I L I T Y S TAT E M E N T
The authors confirm that the data supporting the findings of this study are available within the article [and/or]