Interventions to enhance safety culture for nursing professionals in long-term care: A systematic review

Abstract

Background

Recognizing safety risks and promoting safe care is essential for care dependent people and should be an integral part of the overall preventive endeavors while providing care. The term 'safety culture' describes efforts regarding the acknowledgement and reduction of safety risks. Enhanced safety culture in health care organizations can be associated with a lower incidence of missed nursing care and adverse events.

Objective

Identify strategies to enhance safety culture in long-term care settings and describe factors facilitating or inhibiting the process from the available evidence.

Design

This systematic review is a narrative description of intervention studies.

Methods

CINAHL and MEDLINE were searched in May 2020 using terms such as safety culture, safety climate, intervention development. Gray literature was searched between May and September 2020. An additional search in Cochrane Library was conducted in September 2022. Only intervention studies feasible to enhance the safety culture were included. Intervention study criteria were met when an organized, planned action to prevent or change a specific behavior took place and when this action was transparently and systematically evaluated. The screening, data extraction, and rating processes were conducted by two researchers independently. The ROBINS-I tool was utilized to assess the risk of bias of the studies.

Results

Seven intervention studies were included, all evaluated with a critical risk of bias. Strategies found to enhance the safety culture in care settings include collegial exchange of experiences and learnings, integration of staff's perceptions, external facilitation, staff training, and a structured, multi-step procedure of the intervention process. Some studies were unable to show statistically significant enhancement in safety culture from the interventions implemented. Factors facilitating the implementation of interventions include good connections and trust between staff and managers, and the manager's active support of the project goals, as well as targeting achievable ideas considering time and resources. Time pressure, heavy workloads and high staff turnover may inhibit the process.

Conclusions

All included studies had a high risk of bias, and possible effects must be considered accordingly. Overall, there was considerable heterogeneity in interventions aiming to enhance safety culture. Despite these aspects, promising approaches are training staff's knowledge and competencies regarding open communication and teamwork as part of a multifaceted program. Future research would benefit from participative, carefully developed, comprehensively evaluated interventions for enhancing safety culture, specifically within in-home care settings.

Registration

The review was not pre-registered but described on the website of the Center for Quality in Care.

Tweetable abstract

Participatory change management & staff training help ensure momentum & trust in endeavours to enhance safety culture in long-term care.

Contribution of the paper

What is already known?

• -Recognizing safety risks and promoting safe care is essential for care dependent people and should be an integral part of the overall preventive endeavors while providing care.
• -Enhancing the safety culture in health care organizations can be associated with a lower incidence of missed nursing care and adverse events.
• -Nurse managers and their teams play a vital role in establishing and maintaining a positive safety culture.

What does this paper add?

• -Interventions to enhance the safety culture in long-term care settings are poorly explored, particularly in the in-home care setting.
• -In some studies, participating in comprehensive efforts to enhance safety culture was not only associated with positive changes regarding safety culture but could also improve specific quality and safety indicators such as infection rates, medication safety, falls, or occurrence of pressure ulcers for the care recipient.
• -High staff retention rates provide stability, ensure momentum in implementing changes and sustain trust both within networks and outside the care facility when implementing interventions to enhance safety culture.

1. Introduction

The number of people in need of care is predicted to increase until 2060 as life expectancy rises in Europe (Social Protection Commitee (SPC) and European Commission (DG EMPL), 2021). People in need of care are vulnerable to numerous health risks during long-term care, e.g., medication, nutrition, agility, falls, infections, and even violence (Lachs and Pillemer, 2015; Roquebert et al., 2021; Schreiber et al., 2018; Sloane et al., 2008; Storms et al., 2017). Thus, recognizing safety risks and promoting safe care is essential for care dependent people and should be an integral part of the overall preventive endeavors while providing care (World Health Organization (WHO), 2021). Efforts regarding the acknowledgement and reduction of safety risks are usually summarized in the term 'safety culture', a concept that originally emerged after the nuclear disaster in Chernobyl in 1988. The term was adopted by the health sector following the report ’To err is human’ (Kohn et al., 2000) that sparked a paradigm shift in the contemplation of errors and patient safety. Since various factors (e.g. leadership, communication) are part of safety culture, there is not yet an universal definition for it (O'Donovan et al., 2019). A common description refers to safety culture as being the product of individual and group-related attitudes, beliefs, values, competencies, and behaviors with which members of an organization are committed to safety-related problems (American College of Healthcare Executives, 2017; Cooper, 2000; Guldenmund, 2000; Pfaff et al., 2009). For example, openly facing mistakes and a willingness to learn from them are aspects of a good safety culture.

Theoretically, safety culture is one aspect of the overarching organizational culture of a facility in charge of delivering care. The concept by Schein summarizes organizational culture under the three levels i) artifacts, ii) espoused beliefs and values, as well as iii) underlying assumptions (Gartshore et al., 2017; Schein, 2004). The theoretical underpinnings indicate that safety culture – within the organizational culture – consists of different dimensions worth considering.

Enhancing the safety culture in health care organizations can be associated with a lower incidence of missed nursing care (Hessels et al., 2019) and adverse events (Han et al., 2020; Hessels et al., 2019). Therefore, safety culture can also contribute to the quality of care. Arrangements of quality assurance such as evidence-based practice can make a statistically significant contribution to the safety of people receiving care (Cullen et al., 2020; Gillespie et al., 2016).

Nurse managers and their teams play a vital role in establishing and maintaining a positive safety culture (Han et al., 2020; Hickner et al., 2016; Manser et al., 2016). However, there is still little evidence on the effectiveness of interventions promoting safety culture. Beyond that, interventions have not been sufficiently and thoroughly reported to enable their repeatability and comparability.

Examples of common research topics in clinical settings include ambitions and interventions to avoid medical errors and patient harm. Implementing a critical incident reporting systems (CIRS) is one example for such interventions (Gartmeier et al., 2017; Petschnig and Haslinger-Baumann, 2017). However, little is known about the use and effectiveness of interventions and tools to enhance the safety culture in long-term care settings (Halligan and Zecevic, 2011). Considering both nursing homes and in-home care as long-term care settings, this is especially true for the latter (Bienassis and Klazinga, 2022; Oliveira et al., 2020). Professional nurses in the long-term care settings are incumbent upon complex tasks and responsibilities like provision and monitoring of health care, care coordination as well as supervision and evaluation of the care process (Organisation for Economic Co-operation and Development (OECD), 2020). It can be assumed that they are therefore in a unique position regarding the safety of long-term care facilities.

This systematic review aims to identify strategies to enhance safety culture in long-term care settings and describes factors facilitating or inhibiting the process from the available evidence.

The research questions are, therefore:

• 1What strategies have been used to enhance safety culture in long-term care settings for nurses?
• 2What are the factors facilitating and inhibiting safety culture enhancing interventions in organizations?

2. Methods

This systematic review was part of a comprehensive project on safety culture in the in-home care setting including a scoping review, interviews, and development of educational material for home care agencies on that matter. This systematic review was derived to go into more depth about targeted interventions on safety culture in care settings. The related project was not pre-registered but described on the Center for Quality in Care's website.

2.1. Searching process

The search strategy was developed by two researchers identifying keywords as well as MeSH terms and CINAHL subject headings where applicable based on relevant literature on safety culture. To validate the search strategy and optimize for quantity and quality of results, it was pretested using unique indexing terms and search strategies for MEDLINE and CINAHL. A rather narrowly defined search with subtopics was selected to maximize the relevance of publications and reduce the number of irrelevant studies that did not meet the inclusion criteria. MeSH terms were combined with subheadings when it increased the specificity of the results in the pretest. MEDLINE and CINAHL were searched by using EBSCOhost in May 2022. An additional search was conducted in September 2022 in the Cochrane Library. The finite search strategy for the databases MEDLINE, CINAHL and Cochrane Library consisted of three pillars. The first pillar identified publications about safety culture, including related terms such as patient safety, risk management, and nursing errors. The second pillar identified publications related to the setting using terms such as long-term care, home care agencies and nursing home. The third pillar identified what was summarized by the researchers as intra-organizational process, including terms like nursing process, intervention development and change management. For a more specific search in MEDLINE, a fourth pillar was added focusing particularly on the population. This supplement was waived in the search strings for the other databases as it reduced the number of results in the pretest too much. For the other databases, the limitation of the population took place manually during the screening process.

For the additional hand search of gray literature, journals in German not represented by the databases were prescreened for relevance and validated with the assistance of a gerontological librarian. The hand search was then conducted between May and September 2020. Due to the dearth of literature around interventions to enhance the overall safety culture, a forward citation tracking of studies previously included was performed to find additional literature (Hirt et al., 2020). Regarding the need for the latest evidence on the topic, only references from 2015 onwards were tracked, and citations between 2018 and 2021 have been included when coherent with the inclusion criteria. Tracking was performed via Google Scholar® because there were more distinguishing filter options compared to MEDLINE and CINAHL.

Table 1 provides the MEDLINE search string. Similar search strings – without specifications on population – were used for the other databases. Search-terms and MeSH-terms are in bold and the combined subheadings in italics.

Table 1.

Operators for literature search in MEDLINE, subheadings presented in italics.

Safety culture	patient safety [MeSH]/organization and administration/standards OR safety culture OR risk management [MeSH]/methods/organization and administration/standards/trends OR error management OR quality of health care [MeSH]/adverse effects/education/instrumentation/methods/nursing/organization and administration/prevention and control/standards/surgery/therapy OR safety climate OR safety measures OR quality challenges OR medical errors [MeSH]/adverse effects/instrumentation/methods/nursing/organization and administration/prevention and control/standards/therapy/trends OR nursing error OR adverse event
AND
setting	ambulatory care [MeSH]/adverse effects/education/instrumentation/methods/nursing/organization and administration/pharmacology/rehabilitation/standards/therapy/trends OR outpatient care OR home care agencies [MeSH]/organization and administration/standards/trends OR home care services [MeSH]/adverse effects/education/instrumentation/methods/nursing/organization and administration/pharmacology/rehabilitation/standards/therapy/trends OR nursing [MeSH]/education/instrumentation/methods/organization and administration/standards/therapy/trends OR long-term care [MeSH]/education/ methods/nursing/organization and administration/standards/therapy/trends OR nursing care [MeSH]/education/instrumentation/methods/organization and administration/standards/trends
AND
intra-organizational process	nursing process [MeSH]/education/instrumentation/methods/organization and administration/standards/trends OR intervention development OR change management [MeSH]
AND
population	nurses [MeSH]/education/methods/organization and administration/standards/ therapy/trends OR nursing services [MeSH]/education/methods/organization and administration/standards/trends OR nursing services [MeSH]/education/methods/ organization and administration/standards/trends OR quality management officer OR outpatients [MeSH]/education/organization and administration OR inpatients [MeSH]/education/injuries/organization and administration/rehabilitation/therapy

Inclusion criteria

• ■Types of participants: managers, front staff, clients/patients/habitants ages 18 years and above
• ■Setting: in-home care, nursing home care, clinical care (including all care provided in hospital settings, both acute and rehabilitative)
• ■Target outcome: enhancements in safety culture
• ■Searching period: 2000 – 2020
• ■Language: German & English
• ■Status: completed
• ■Design: intervention study

If studies included nurses and other professions as a team, they were only included in this systematic review when results for nurses were collected, analyzed, and presented independently from those of other professions.

Studies were excluded if they focused on non-nursing professionals like relatives of care dependent people, physicians or therapists; were set in intensive, palliative or pediatric care; were published before 2000; were incomplete or not finished; were non-intervention studies or targeted only patient or resident related outcomes without measuring safety culture outcomes in addition.

Intervention studies on safety culture specifically in nursing homes and in-home care were limited. It was therefore decided during the pretest of the search syntax to not exclude studies conducted in the clinical setting if they focused on safety culture for nurses. However, studies conducted in the clinical setting were not explicitly searched for.

As the focus of this systematic review was based on interventions feasible to enhance the safety culture, the definition for intervention studies was defined as organized, planned action to prevent or change a specific behavior and a transparent and systematic evaluation of this action. These criteria were added to the inclusion/exclusion criteria accordingly.

‘Enhancing’ of the safety culture was determined by whether the safety culture was assessed in a comprehensible manner using quantifiable or qualitative parameters, e.g., using specific instruments or the content analysis of qualitative interviews with regards to the intervention carried out.

2.2. Review process

The review process was conducted by independently screening titles, abstracts, and full texts by two reviewers. Selections were discussed at every stage of the revision process. In case of disagreement, a third reviewer decided to include or exclude the paper in question with precise regard to the inclusion and exclusion criteria. A review protocol and data extraction protocols used are available from the first author upon request.

2.2.1. Data extraction

A data extraction protocol was prepared and used for each study. Extractions were performed by one person per study, summarizing methods and results. No author was contacted for additional study information. The original plan of conducting a meta-analysis was rejected due to insufficient data. Therefore, the outcomes of the intervention studies were processed for a systematic review as a narrative summary of interventions feasible to enhance safety culture for long-term care settings.

2.2.2. Assessment of methodology

The ROBINS-I tool was utilized to evaluate the methodology of non-randomized study designs (Schünemann et al., 2019; Sterne et al., 2016). The overall risk of bias for a study assessed with the ROBINS-I can only be rated as high as the domain with the lowest rating. Different ratings range on a scale from ‘critical’_’, ’serious’, ‘moderate‘ to ’low‘ risk of bias, or no information, respectively. All studies were rated following the guideline procedure for ROBINS-I by two authors independently. Disagreements were discussed between both authors. In case there was no consent, a third author decided with regard to the ROBINS-I tool requirements.

It was planned to use the RoB-2 tool for randomized studies (Sterne et al., 2019), but it has not been proven necessary since no randomized controlled trials (RCTs) met the inclusion criteria.

3. Results

3.1. Study selection

Fig. 1 shows the records flow for the literature search, screening, and remaining papers after excluding non-relevant studies for this review.

Fig. 1.

Flowchart of the study selection.

3.2. Study characteristics

Total data extraction presented in Table 2 gives an overview of the characteristics of all included studies. The seven included papers were conducted in England (n = 2), United States of America (USA) (n = 2), Canada, Germany, and Norway. No intervention studies on safety culture were found in the in-home care setting. Interventions addressed enhancing the organizations' safety culture as per inclusion criteria. However, they presented various additional outcome parameters and approaches resulting in different data assessment methods. No RCT matched the inclusion criteria.

Table 2.

Data extraction table for all included intervention studies.

First author (year)	Country	Study purpose	Design	Assessment method	Participants	Facility sample size, Setting	Duration	Theory underpinning & Intervention	Main outcome
Agency for Healthcare Research and Quality (AHRQ) (2017)	USA	Evaluation of a safety program development and implementation to reduce healthcare-associated infections (HAI), especially catheter catheter-associated infections (CAUTI)	Data collection during the intervention, no control group	Standardized assessment of facility demographics, Nursing Home Survey on Patient Safety Culture (NHSPSC), infection rates, process data, qualitative interviews	1 Nurse champion, 1 physician champion, 1 infection preventionist, key members of the clinical staff (e.g., registered nurses, licensed practical nurses, certified nursing assistants	n = 505 Long term care facilities	10 - 12 months per cohort	No underpinning theory. Comprehensive Unit-Based Safety Program for long term care utilizing evidence-based infection prevention focusing on the reduction of catheter use; intervention with a focus on good safety culture, training on teamwork, and communication (TeamSTEPPS®)	Aggregated (per definition) as well as population-based infection rates decreased statistically significantly (95% CI 0,42–0,63), slight increase in 8/12 safety culture domains
Damery et al. (2019)	ENG	Evaluation of the implementation, staff experiences, and implemented safety interventions of the safer provision and caring excellence (SPACE) program and its effects on preventing avoidable harm, resident safety outcomes, and safety climate	Pre-post intervention, no control group	Safety Attitudes Questionnaire (SAQ), standardized assessment of facility demographics, second data analysis of routine data to adverse events, semi-structured interviews, team observations during training and key meetings	Care home managers, non-manager care home staff	n = 29 Care homes	24 months	Appreciative inquiry as underlying principle/framework: emphasizing what is going well and applying the learnings. Skill training in quality improvement techniques like the use of generic instruments for risk assessments, tools to improve communication for managers and nurses, utilization of data to support quality improvements, empowering staff to verbalize and execute ideas and shared learnings	High variance, but statistically significant reduction in fall rates (p = 0.039 in one region, p = 0.001 in the other region), partially statistically significant reduction in urinary tract infections (p = 0.001 in one region & p = 0.34 in the other region), insignificant increase in pressure ulcers and hospital admissions, improvements in all SAQ-dimensions that were associated with working full time, higher qualification, participation in training, smaller facility size
Ganaden and Mitchell (2018)	CAN	Evaluation of program implementation to enhance safety culture and medication safety	Pre-post intervention, no control group	Canadian patient safety climate survey (Can-PSCS)	22 Care staff members of the facility – 18 health care aides, 1 health care aides supervisory resource coordinator, 2 clients case coordinators	n = 1 Assisted living facility	4 months	Lean Six Sigma improvement activities including affinity diagram exercise, job shadowing and current state process mapping as underlying framework. 8-step Comprehensive Unit-Based Safety Program (CUSP) including pre-assessment, education, developing an action plan based on assessment, leadership support, implementation, evaluation, sharing stories & documentation	Self-reported increase in perceptions of patient safety, learnings, and occurrence reporting
Mai (2015)	GER	Assess to what extent collegial case advice meetings can raise collective awareness and can therefore enhance a culture of learning and communication	Controlled intervention	Safety Organizing Scale (SOS), qualitative observation, and group discussions	4–5 Nurses participating on average per collegial case advice meeting (19 collegial case meetings in total)	n = 1 Acute hospital	6 months	No underpinning theory. Structured and moderated collegial case advice meetings based on a checklist with a focus on fall prevention, subsequent reflection talks, and discussions on different perspectives in feedback rounds	No statistically significant improvement on the Safety Organizing Scale; awareness for the prevention of falls was reported to be increased
Marshall et al. (2018)	ENG	Evaluation of a complex safety improvement intervention and its impact on care home culture and work processes	Pre-post intervention, no control group	Qualitative interviews, observations, documentation reviews, simple online survey based on Manchester Patient Safety Framework	Managers and front-line staff of the care homes, non-care home stakeholders, including health service staff, social- and healthcare commissioners, small number of family members and residents	n = 90 Residential and nursing care homes	6 months	Theory: Promoting safer provision of care for elderly (PROSPER) program to lower the incidence of falls, pressure ulcers, and urinary tract infections by training of quality improvement methods, train-the-trainer-approach, and utilization of the Manchester Patient Safety Framework; enabling staff to evaluate resident related outcomes	Qualitative evidence that the program has led to notable changes in the perception of safety culture and working practices
Smith et al. (2018)*	USA	Test the hypothesis that nursing homes with higher initial safety culture or larger enhancements in safety culture have greater success in preventing catheter-associated urinary tract infections (CAUTI)	Secondary data analysis	Nursing Home Survey on Patient Safety Culture (NHSPSC), CAUTI rates	Nursing home staff (10 or more per nursing home)	n = 196 Community-based nursing homes	10 - 12 months per cohort	No underpinning theory. Implementation of technical (professional training in the use of catheters) and socio-adaptive components (leadership engagement and team communication) to enhance structure, process, and practice of infection prevention and to enhance safety culture	52% Reduction in CAUTI-rates, statistically insignificant improvements as well as declines in different dimensions of the NHSPSC, no statistically significant association between facility-level safety culture and rates of catheter-associated urinary tract infections
Storm et al. (2018)	NOR	Assess the effects of inter-organizational educational staff meetings with a focus on quality and safety in transitional care and safety culture	Quasi-experimental, pre-post intervention	Hospital Survey on Patient Safety Culture (HSOPSC), qualitative data from follow-up meetings on initiated measures and organizational changes	Nurses, nursing assistants, medical doctors, ward leaders	n = 1 City based university hospital (5 wards, pulmonary medicine and emergency medicine in intervention group)	5 months	No underpinning theory. Inter-organizational educational intervention program ’Meeting Point’ as half-day seminars including educational sessions and a discussion platform with participants from different professions and settings (hospital, nursing home, home care services, patient coordination office) with topics around improving transitional care; risk factors, patient perspectives, and system perspectives	Small but statistically significant (95% CI 0.19 to 0.80) change in overall perceptions of safety culture immediately after intervention and after 12 months of follow-up. Improved handoffs & transition, organizational learning, and overall perceptions of safety culture in the intervention group only; however, concurrent implementation of an electronic patient flow registration system not connected to the program could have positively impacted these outcomes
				Nursing Home Survey on Patient Safety Culture (NHSPSC), qualitative data from follow-up meetings on initiated measures and organizational changes	Nurses, nursing assistants, medical doctors, ward leaders	n = 3 Municipality nursing homes (3 wards, intermediate length of stay in intervention group)			Negative development in most patient safety culture dimensions over time; however, unexpected organizational changes not linked to the program may have impacted these results

^⁎Secondary data analysis of the AHRQ program (2017), but with a reduced sample size and slightly different focus.

3.3. Risk of bias

Results of the ROBINS-I tool for the studies included are found in Fig. 2. The vast majority of intervention studies were non-randomized, unblinded, without a control group, and had not accounted for potential confounders on the care recipient side (e.g., age or diseases) or potential confounders of the caregiver side (e.g., work experience or age). In addition, some studies were conducted single-centered and with a relatively small number of participants resulting in low effect power and, therefore, low transferability. Consequently, all studies evaluated had ’critical‘ risks of bias. Full rating assessments are available upon request from the first author.

Fig. 2.

Risk of bias for each study visualized by using the robvis tool (https://www.riskofbias.info/welcome/robvis-visualization-tool).

3.4. Strategies to enhance safety culture in long-term care settings

3.4.1. Collegial exchange of experiences and learnings

A growing culture of sharing information and learning from shared experiences when attempting to enhance an organization's safety culture were recurring topics in six of seven studies (Agency for Healthcare Research and Quality (AHRQ), 2017; Damery et al., 2019; Ganaden and Mitchell, 2018; Mai, 2015; Marshall et al., 2018; Storm et al., 2018). During the two-year safer provision and caring excellence (SPACE) program in England, shared learning experiences resulted in learning from each other and improved teamwork. In addition, it was noted that the regular feedback helped develop a change process toward quality improvement and safety culture within the organization (Damery et al., 2019). That is in line with the findings of a study based in an assisted living facility in Canada. Researchers found that getting team feedback on nurses' experiences was described as an essential resource for the staff's work orientation. In the same program, organizational learning was supported by employees writing down questions often verbalized by clients regarding their medication. The questions collected were used for group discussions resulting in enhanced team collaboration (Ganaden and Mitchell, 2018). The benefits of learning in groups were described as well in a six-month program promoting safer provision of care for elderly residents (PROSPER), which included ninety nursing homes in England where experiences were exchanged not only within the ward team but in networking and meetings with other nursing facilities. Participating in the program led to nurses managing residents' risks more proactively and better reflection on nurses' role regarding safety and improvements in daily workflows (Marshall et al., 2018). Positive effects of sharing knowledge and experiences were also seen among stakeholders (i.e., residents and their family members, facility staff, associated organizations) in a sizeable US-wide study that implemented a program to reduce catheter-associated urinary tract infections by integrating team strategies and tools to enhance performance and patient safety (TeamSTEPPS®) training and communication (Agency for Healthcare Research and Quality (AHRQ), 2017).

Two studies with elements of experience exchange as part of the intervention to enhance safety culture did not show precise positive results: In a Norwegian study, interprofessional meetings with other stakeholders, including educational sessions and a discussion platform focusing on quality and safety in transitional care showed slight enhancements in some safety culture factors assessed for the hospital group. The authors concluded that these positive developments could not be solely connected to the intervention as the hospital introduced an electronic patient flow registration system during the intervention period. There were no effects on safety culture in nursing homes that received the same intervention. However, organizational changes unrelated to the intervention could have played a factor in that outcome(Storm et al., 2018). In a German study, structured collegial case advice meetings in a hospital ward did not lead to higher ratings of the safety culture compared to a control group (Mai, 2015).

3.4.2. Integration of staff's perceptions

Three programs addressed the focus of the interventions executed on previously assessed needs of staff (Damery et al., 2019; Ganaden and Mitchell, 2018; Marshall et al., 2018). During the Comprehensive Unit-based Safety Program (CUSP), nurses in an assisted living facility in Canada initially identified safety concerns with the Canadian Patient Safety Culture Survey Tool (Can-PSC). Subsequently, they developed and agreed to an action plan using the Cultural Check-up Process Tool published by the Agency for Healthcare Research and Quality (AHRQ) (Ganaden and Mitchell, 2018). A participative method was also used in the Promoting safer provision of care for elderly (PROSPER) program in England, where the employees of nursing homes developed and accompanied the complex intervention implemented. The nurses reported that their involvement in the process raised the awareness of their role in creating a safer environment (Marshall et al., 2018). Similar effects were seen during the safer provision and caring excellence (SPACE) program in nursing homes in England: Letting nurses co-create improvements instead of imposing standardized procedures enabled them to express and execute their ideas according to the authors (Damery et al., 2019). All three studies described positive changes toward safer daily workflows and aspects of safety culture.

3.4.3. External facilitation

The intervention process was supported by external facilitators in all studies included. Facilitators were, e.g., moderators from the research team (Mai, 2015) or backed through the organizational network of the program (Agency for Healthcare Research and Quality (AHRQ), 2017). Bringing in new methods, tools, and expertise was described as beneficial for mobilizing nurse's commitment to the program (Marshall et al., 2018), understanding how to make surveillance data actionable, and identifying areas with room for improvement (Agency for Healthcare Research and Quality (AHRQ), 2017; Damery et al., 2019).

3.4.4. Staff training

All studies described educational elements during or in preparation of the intervention that targeted improvement of knowledge on safety-related topics and communicational aspects – however, the context of these educational sessions was diverse. In efforts to improve transitional care in hospitals and nursing homes, training units were held on transitional care as frequent day seminars with evolving content and feedback from the participants. These training units were organized by the research team with participants from different care settings (hospitals, nursing homes, in-home care, patient coordination offices) (Storm et al., 2018). During the Comprehensive Unit-based Safety Program (CUSP) in an assisted living facility in Canada, nurses received a 40-minute presentation with different theoretical inputs on safety culture. These included Reason's theory on human error and systems-based thinking (Reason, 2000), speaking-up and listening, and lectures on the acknowledgement of personal and organizational vulnerabilities as well as accepting responsibility without blaming one another (Ganaden and Mitchell, 2018). Two programs utilized a train-the-trainer approach, in which trained persons shared educational content on quality improvement and safety culture knowledge with the frontline nurses (Agency for Healthcare Research and Quality (AHRQ), 2017; Marshall et al., 2018). Damery et al. (2019) described various and ongoing harm-themed training that was held both in nursing homes and as centrally organized sessions. Training content was flexible and individualized, using the language and examples of the specific facilities and aimed to combine theory and practicality. Participants of some studies stated that they liked the training itself, especially learning in groups (Agency for Healthcare Research and Quality (AHRQ), 2017; Damery et al., 2019; Marshall et al., 2018). Others showed positive effects in the organizational learning domains of the safety culture assessment (Agency for Healthcare Research and Quality (AHRQ), 2017; Ganaden and Mitchell, 2018; Storm et al., 2018). Nevertheless, in some evaluations, the organizational learning domain slightly but statistically significant decreased throughout the intervention (Agency for Healthcare Research and Quality (AHRQ), 2017; Smith et al., 2018). The provision of various training to cover different learning styles was perceived positively, and managers (Damery et al., 2019) as well as nurses (Ganaden and Mitchell, 2018), reported to have learned helpful information for future application during training sessions. Additionally, it was evident from quantitative and qualitative surveys that skill training led to 'specific improvements to multiple areas of safety within participating care homes' (Damery et al., 2019). One study did not assess and report the learning effects of educational elements (Mai, 2015).

3.4.5. Structured, multi-step procedure

Most of the interventions implemented in the studies aiming to enhance safety culture were based on an extensive methodology with several different elements. Surveys on safety culture or safety-related outcomes, e.g., were used to identify areas of improvement for which targeted intervention strategies (see above) were subsequently implemented (Damery et al., 2019; Ganaden and Mitchell, 2018; Marshall et al., 2018; Storm et al., 2018). Of these, three enabled the nurses to use their own data to evaluate the targeted safety outcomes (Damery et al., 2019; Ganaden and Mitchell, 2018; Marshall et al., 2018). Marshall et al. (2018) as well as Damery et al. (2019), introduced quality improvement techniques like the Plan-Do-Study-Act (PDSA) cycles for that purpose. One program had precast technical and social-adaptive components to address specific safety-related outcomes like catheter-associated urinary tract infections and the overall safety culture (Agency for Healthcare Research and Quality (AHRQ), 2017; Smith et al., 2018). Social-adaptive components included team formation, excellent communication, assessing what's working, meeting monthly, and sustaining efforts (Agency for Healthcare Research and Quality (AHRQ), 2017).

3.5. Factors facilitating and inhibiting the implementation of interventions to enhance safety culture

Several factors were identified by the researchers to have had a positive or negative impact on implementing interventions to enhance the organization's safety culture. Three studies describe facilitating and inhibiting factors (Agency for Healthcare Research and Quality (AHRQ), 2017; Damery et al., 2019; Marshall et al., 2018). Furthermore, two studies each describe only facilitating (Ganaden and Mitchell, 2018; Smith et al., 2018) or inhibiting (Mai, 2015; Storm et al., 2018) factors. The results from all included studies can be grouped into three segments: staff (i.e., trust between staff and managers vs. high staff turnover), management (i.e., active support vs. fluctuation), and program organization (i.e., achievable ideas regarding time and resources vs. high preparation efforts in relation to the benefit). Other key factors from these segments are displayed in Table 3.

Table 3.

Facilitating and inhibiting factors for implementing interventions to enhance safety culture.

Facilitating	Inhibiting
Staff
Empower to implement change (Damery et al., 2019; Smith et al., 2018)	Time pressure and heavy workloads and, therefore, competing demands and project fatigue or reduced willingness to complete surveys (Agency for Healthcare Research and Quality (AHRQ), 2017; Mai, 2015; Marshall et al., 2018)
Responsibility to specific areas of safety (Damery et al., 2019)	High staff turnover and staff shortage (Marshall et al., 2018; Storm et al., 2018)
Good connections and trust between staff and managers (Damery et al., 2019)	Feeling of patronage and surveillance (Mai, 2015), mistrust, and skepticism towards the participation (Damery et al., 2019)
Management
Backing and active support of the project goals (Ganaden and Mitchell, 2018; Marshall et al., 2018; Smith et al., 2018)	Fluctuation of managers (Damery et al., 2019; Marshall et al., 2018; Storm et al., 2018)
Program organization
Target achievable ideas considering time and resources available (Ganaden and Mitchell, 2018), flexible intervention approach so facilities can adapt materials to their needs (Agency for Healthcare Research and Quality (AHRQ), 2017)	High preparation effort in relation to the benefit (Mai, 2015)
Integrate changes into the workflow and organizational culture (Damery et al., 2019)	Organizational changes like restructuring of the facility taking place simultaneously with the intervention (Storm et al., 2018)
Offer tailored training to fit nurses' schedules and learning styles, use relevant language and examples (Damery et al., 2019), point out transfer options in strategies learned from other areas of practice (Agency for Healthcare Research and Quality (AHRQ), 2017)
Bring in external expertise and network with other nursing homes (Damery et al., 2019; Marshall et al., 2018), good relationships between organizational leads and facilities (Agency for Healthcare Research and Quality (AHRQ), 2017)
Support through program moderators (Damery et al., 2019; Marshall et al., 2018), ongoing coaching in the desired practices (Agency for Healthcare Research and Quality (AHRQ), 2017)	Inconsistency of the team facilitating the program (Marshall et al., 2018)
Non-judgmental feedback on changes, utilizing own data (Agency for Healthcare Research and Quality (AHRQ), 2017; Damery et al., 2019; Marshall et al., 2018)
In-person meetings (where possible) to forge stronger connections (Agency for Healthcare Research and Quality (AHRQ), 2017)
Transparent processes regarding occurrence reporting and the option of reporting anonymously (Ganaden and Mitchell, 2018)	Technical difficulties like lack of computers with internet (Agency for Healthcare Research and Quality (AHRQ), 2017)
	Insufficient submission of assessments, so that feedback is only possible to a limited extent (Agency for Healthcare Research and Quality (AHRQ), 2017)

4. Discussion

This paper aimed to provide a systematic review of safety culture interventions in long-term care settings and analyze the evidence. The international systematic literature search ultimately revealed only seven studies that could be included in the analysis. The studies took part in different countries, mainly in nursing homes or clinical settings. This is in line with a previous scoping review that reported a lack of research on safety culture for the in-home care setting (Gartshore et al., 2017). This review showed a very heterogeneous picture of intervention approaches. Most interventions addressed a mixture of different components, which reduces their comparability. Overall, the studies included have a high risk of bias and thus should be interpreted cautiously. Individual aspects, as well as the methodology will be discussed in the following sections.

Several promising intervention strategies regarding the research question were identified. Main approaches include the collegial exchange of experiences and learnings, integration of staff perceptions, external facilitation, and staff training embedded in a structured, multi-step procedure. Having a constructive and transparent process in place where incidents were reported orderly helped to create an environment of reporting and learning from mistakes (Ganaden and Mitchell, 2018). Some of the results indicate that participating in comprehensive efforts to enhance safety culture was not only associated with positive changes regarding safety culture but could also enhance specific quality and safety indicators such as infection rates (Agency for Healthcare Research and Quality (AHRQ), 2017; Smith et al., 2018), medication safety (Ganaden and Mitchell, 2018), falls or occurrence of pressure ulcers for the care recipient (Damery et al., 2019). Nevertheless, four research teams could not prove positive effects of the executed intervention on the overall safety culture for the facilities in question (Agency for Healthcare Research and Quality (AHRQ), 2017; Mai, 2015; Smith et al., 2018; Storm et al., 2018). Besides the possibility of the particular intervention not being able to change the safety culture for the better, there could have been other factors contributing to this outcome. According to the respective authors, this could, e.g., be due to inconsistency in the facility and organizational structures (Storm et al., 2018), reservations towards the approach of the intervention (Mai, 2015), or the chance that the instrument used to assess safety culture was unable to capture specific, individual changes (Smith et al., 2018). Still, longer-term evaluations and larger sample sizes are needed to draw more reliable conclusions – specifically when measuring the complex dynamics of safety culture within an organization. Safety culture may require three to five years for observable, measurable changes (Connor et al., 2007; Halligan and Zecevic, 2011; Smith et al., 2018).

Steady staffing of nurses and managers was reported to play an important role in not losing the momentum in the change process, and it showed to be important to sustain trust and networks within and outside the care facility (Agency for Healthcare Research and Quality (AHRQ), 2017; Damery et al., 2019; Marshall et al., 2018). The empowerment of nursing staff to implement change and having the project goals backed and supported by the management facilitated program efforts (Damery et al., 2019; Ganaden and Mitchell, 2018; Marshall et al., 2018; Smith et al., 2018). This is in line with further findings from other studies, showing that creating a positive learning environment as well as encouraging nurse leaders to facilitate knowledge management might be beneficial for care recipient outcomes (Halligan and Zecevic, 2011; Lunden et al., 2017; Wong et al., 2013). Similar hints about the value of strengthening communication skills come from three previously published reviews which verified positive effects on safety culture through mindfulness and improved communication by leaders (Roussel, 2019), effective interprofessional communication (O'Donovan et al., 2019), and nurse leaders supporting the process of acquiring, sharing, and creating new knowledge (Lunden et al., 2017). The present results are also supported by a recent review from the clinical setting. It reported that, e.g., knowledge improvement, collaboration in performing tasks, regular feedback, and electronic systems for communication could help enhance adherence to patient safety principles like risk management, infection control, or providing a safe environment (Vaismoradi et al., 2020).

On a theoretical level, safety culture was described as a construct built on all employees' shared attitudes, beliefs, values, competencies, and behaviors (Cooper, 2000; Guldenmund, 2000; Pfaff et al., 2009). This complex understanding also makes safety culture challenging to measure, as staff may have limited knowledge, exposure to, or influence over the aspects of culture (Smith et al., 2018). In addition, many studies still lack a shared understanding and concept of safety culture in the demarcation of the term safety climate (Cooper, 2000; Gershon et al., 2004; Guldenmund, 2000). Safety climate is often defined as 'features of the safety culture from attitudes and perceptions of individuals at a given point in time' (Flin et al., 2000). This demarcation into safety culture and safety climate would mean that the time component must be taken into account in the context of a study, e.g., whether a point in time or a longer period of time is being considered. This would also support the assumption that safety culture can be viewed as a continuum, being subject to temporal fluctuations, whereas safety climate is fixed to one point in time. In order not to confuse this temporal dependency, it is important for future studies to be precise about the use of terminology. That variance was found in the studies included in this review as well. Even though all studies used instruments or qualitative interviews to assess the safety culture onsite, only Damery et al. (2019) referred to it as ‘safety climate’ rather than ‘safety culture’. It is possible that the use of terminology effects the orientation and expectation of interventions to promote change.

As safety culture is much more than the sum of different safety measurements for the person receiving care, corresponding individual factors are not always easy to evaluate. To assess whether safety culture enhanced from an intervention implemented is therefore highly complex. One possibility is to use questionnaires that have been validated for clinical (e.g., HSOPSC) and nursing home (e.g., NHSOPSC) settings (Gartshore et al., 2017; Sorra et al., 2018; Sorra and Dyer, 2010). However, validation studies showed substantial differences in the underlying dimensions and psychometric properties in different countries (Cappelen et al., 2016; Hedsköld et al., 2013; Perneger et al., 2014; Pfeiffer and Manser, 2010; Zhao et al., 2019) and an established questionnaire for the in-home care setting is still missing. Using different instruments to assess organizations' safety culture in the studies makes it difficult to compare the outcomes. This, in turn, leads to how to evaluate meaningful changes in safety culture so that intervention effects can be assessed (Halligan and Zecevic, 2011). The intervention studies from the present review used different data collection methods: questionnaires, clinical outcomes, and qualitative interviews about implementation processes and practical adaptations. Thus, using more than one method seems reasonable to evaluate a highly complex construct such as safety culture.

Moreover, qualitative approaches might enable more profound insights into safety culture (Gartshore et al., 2017). Canadian authors also criticized a lack of a common concept and understanding of safety culture in theory as well as in intervention studies (Halligan and Zecevic, 2011) which could be confirmed by analyzing data for the present review. An idea already being used in certain areas of medicine and nursing is measuring core outcome sets (Williamson et al., 2017); selecting appropriate core outcomes for research studies to better compare the effects of interventions. This could potentially reduce heterogeneity and allow better comparisons in future safety culture studies. Another interesting approach to increasing intervention studies' comparability comes from behavioral medicine. The behavior change techniques taxonomy (BCTT) attempts to make group intervention techniques comparable (Michie et al., 2013). In this approach, experts from different disciplines have agreed on technical standards that can be used for various interventions. Regarding safety culture it might be interesting to use standardized assessments (i.e., surveys) and comparable techniques in interventions (i.e., Action Planning, Goal Setting, or Feedback on behavior as adapted examples from above mentioned behavior change techniques).

5. Limitations of this review

There are some methodical issues to discuss. Despite an extensive search in three scientific databases (MEDLINE, CINAHL, Cochrane Library), a hand search, and citation tracking focusing on English and German language, there may be relevant work missing. On the other hand, the databases are commonly used, cover many scientific fields, and are deemed the most comprehensive in health sciences (Tacconelli, 2010). Besides the limitation through the number of databases searched, the search string itself might have limited the quantity of studies found on the subject matter. In addition, we cannot rule out the possibility that more relevant literature could have been found with a more sensitive search. Regarding the focus of this systematic review, the scope was narrowed to only the nursing profession, excluding other professions that provide long-term care. We concluded that researching more precisely for nurses would result in concrete indications for this profession considering that there is limited evidence about ways to enhance the safety culture in their particular work setting. Apart from these search-related concerns, methodical issues are important to consider. Due to the scarce literature on the topic, criteria for inclusion were chosen rather openly, even when that meant reducing specificity. For example, although the review's focus was on the long-term and not the acute care setting, studies set in hospitals were not systematically excluded during screening. This and other methodical, regulatory (e.g., country-specific conditions and costs, staffing characteristics, authorizations, education), setting-related differences (nursing home, assisted living, hospital), various intervention components, and diverse measurement methods impede conclusions and transferability considerably. The resulting heterogeneous study situation made it difficult to reach a firm conclusion regarding the evidence.

Consequently, the authors of this systematic review abstained from summarizing complex intervention studies as this may be critical because of potential bias (Mühlhauser et al., 2011). Thus, heterogenetic studies were evaluated and prepared as narrative summaries. The low methodological quality and the high risk of bias of the studies included could cause bias in the conclusion. To avoid this, conclusions were formulated cautiously.

6. Practical implications

Findings of the systematic review suggest that interventions could represent an additional burden with increased bureaucracy and workload for the nursing team, especially in the beginning. To counteract potential skepticism, it makes sense to have complementary strategies for competing demands in the day-to-day work as well as positive reasoning to the benefits of safety culture in place. When implementing such interventions, it is important to address all employees regardless of their position. Management support, focus on internal and external communication skills, constructive incident management, sufficient workforce and qualifications seem to be essential prerequisites for the success of an intervention to enhance safety culture. A participatory approach that considers the needs of employees is advantageous, too. Examples from the literature with established, structured methods include the Comprehensive Unit-based Safety Program (CUSP), TeamSTEPPS® and the Manchester Patient Safety Framework. Corresponding interventions to enhance the safety culture have been described in the literature with a duration of up to three years.

7. Conclusion

The systematic review presented different interventional approaches to enhance a care facility's safety culture and factors that can potentially facilitate or inhibit the process. The quality of the included studies is low, and possible effects must be considered accordingly. Overall, there was considerable heterogeneity in interventions aiming to enhance safety culture. Despite these aspects, promising approaches are training staff's knowledge and competencies regarding open communication and working together on enhancing safety concerns as part of a multifaceted program (Bisbey et al., 2021). This could include facility data and suitable tools that support learning effects. In future studies, it could be helpful to implement core data sets and standardized techniques to achieve increased comparability. Future research would benefit from participative, carefully developed, and comprehensively evaluated interventions to op safety culture for nursing professionals, specifically in-home care settings. For that reason, the development of a valid quantitative assessment to measure safety culture in-home care/home care agencies is required. Imbed interventions over more extended periods (e.g., three years or more) and regularly evaluating them seems appropriate to reflect the dynamics of safety culture.

Funding

This research did not receive any specific grant from funding agencies in the public, commercial or not-for-profit sectors.

CRediT authorship contribution statement

All authors contributed to the design and conception of the study. SG and MH searched and analyzed the literature and drafted the manuscript. All authors contributed to the interpretation of the results from the literature, revised the manuscript critically, and approved the final version.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.