The role of hospital managers in quality and patient safety: a systematic review

Objectives To review the empirical literature to identify the activities, time spent and engagement of hospital managers in quality of care. Design A systematic review of the literature. Methods A search was carried out on the databases MEDLINE, PSYCHINFO, EMBASE, HMIC. The search strategy covered three facets: management, quality of care and the hospital setting comprising medical subject headings and key terms. Reviewers screened 15 447 titles/abstracts and 423 full texts were checked against inclusion criteria. Data extraction and quality assessment were performed on 19 included articles. Results The majority of studies were set in the USA and investigated Board/senior level management. The most common research designs were interviews and surveys on the perceptions of managerial quality and safety practices. Managerial activities comprised strategy, culture and data-centred activities, such as driving improvement culture and promotion of quality, strategy/goal setting and providing feedback. Significant positive associations with quality included compensation attached to quality, using quality improvement measures and having a Board quality committee. However, there is an inconsistency and inadequate employment of these conditions and actions across the sample hospitals. Conclusions There is some evidence that managers’ time spent and work can influence quality and safety clinical outcomes, processes and performance. However, there is a dearth of empirical studies, further weakened by a lack of objective outcome measures and little examination of actual actions undertaken. We present a model to summarise the conditions and activities that affect quality performance.

• The review reveals conditions and actions conducive to good quality management and offers a model to transparently present these to managers considering their own part in quality and safety.
• The search for this review has screened a vast amount of the literature (over 15,000 articles) across a number of databases.
• The small number of included studies and their varied study aims, design and population samples make generalisations difficult. With more literature on this topic, distinctions could be made between job positions.
• The quality assessment scores are subjective and may not take into consideration factors beyond the quality assessment scale used.  3] In line with this, there have been calls for Boards to take responsibility for quality and safety outcomes. [4,5] One article warned hospital leaders of the dangers of following in the path of bankers falling into recession, constrained by their lack of risk awareness and reluctance to take responsibility. [6] To add to the momentum are some high profile publicity of hospital management failures affecting quality and safety, eliciting strong instruction for managerial leadership for quality at the national level in some countries. [7,8] Beyond healthcare, there is clear evidence of managerial impact on workplace safety. [9][10][11][12] Within the literature on healthcare, there are non-empirical articles providing propositions and descriptions on managerial attitudes and efforts to improve safety and quality. This literature, made up of opinion articles, editorials and single participant experiences, present an array of insightful suggestions and recommendations for actions that hospital managers should take to improve the quality of patient care delivery in their organisation. [13][14][15][16][17] However, researchers have indicated that there is a limited evidence-base on this topic. [18][19][20][21] Others highlight the literature focus on the difficulties of the managers' role and the negative results of poor leadership on quality improvement (QI) rather than considering actions that managers presently undertake on quality and safety. [22,23]   How much time is spent by hospital managers on quality and safety and its improvement?
What are the managerial activities that relate to quality and safety and its improvement?
How are managers engaged in quality and safety and its improvement?
What impact do managers have on quality and safety and its improvement?
How do contextual factors influence the managers' role and impact on quality and safety and its improvement?

Concepts and definitions
Quality of care and patient safety were defined on the basis of widely accepted definitions from the Institute of Medicine [4,24] and literature was searched for all key terms associated with quality and patient safety to produce an all encompassing approach. A manager was  6 defined as any employee that manages staff and is likely to hold managerial responsibilities such as budget responsibilities and staff recruitment and training. Therefore, all levels of managers including Boards of managers were included in this review with the exception of clinical frontline employees, e.g. doctors or nurses, who may have taken on further managerial responsibilities alongside their work but do not have a primary official role as a manager. Those that have specifically taken on a role for quality of care, e.g. the modern matron, were also excluded. Distinction between senior, middle and frontline management were as follows: senior management hold Trust-wide responsibilities;[25] middle managers are in the middle of the organisational hierarchy chart and have one or more managers reporting to them; [26] frontline managers are defined as managers at the first level of the organisational hierarchy chart who have frontline employees reporting to them. Board managers include all members of the Board. Although, there are overlaps between senior managers and Boards (for example Chief Executive Officers (CEOs) may sit on hospital Boards), we aim to present senior and Board level managers separately due to the differences in their responsibilities and position. Only managers that would manage within or govern hospitals were included, with the exclusion of settings that solely served mental health or that comprised solely of non-acute care community services. The definition of 'Role' focused on actual engagement, time spent and activities that do or did occur rather than those recommended that should or could occur.

Search Strategy
Literature was reviewed between 01 Jan 1983 and 01 Nov 2010. Eligible articles were those that described or tested managerial roles pertaining to quality and safety in the hospital setting. Part of the search strategy was based on guidance by . [ ii.Quality; iii.Hospital Setting) and five steps. A facet on role was not included in the search strategy, as it would have significantly reduced the sensitivity of the search.
Multiple iterations and combinations of all search terms were tested to achieve the best level of specificity and sensitivity. In addition to the key terms, Medical Subject Headings (MeSH®) terms were used, which were 'exploded' to include all MeSH subheadings. All databases required slightly different MeSH terms (named Emtree in Embase), therefore four variations of the search strategies were used (see online Appendix 1 for the search strategies). Additional limits placed on the search strategy restricted study subjects to human and the language to English. The search strategy identified 15,447 articles after duplicates had been removed.

Screening
Three reviewers (AP, AR and DG) independently screened the titles and abstracts of the articles for studies that fit the inclusion criteria. One reviewer (AP) screened all 15,447 articles, while two additional reviewers screened 30% of the total sample retrieved from the search strategy: AR screened 20% and DG screened 10%. On testing inter-rater reliability, Cohen's kappa correlations showed low agreement between AR and AP (K=0.157, P<0.01) and between DG and AP (K=0.137, P<0.00). [28] However, there was a high percentage of agreement between raters (95% and 89% respectively), which reveals a good inter-rater reliability. [29,30] Discrepancies were resolved by discussion and consensus. The main inclusion criteria were that: the setting was hospitals; the population sample reported on were managers; the context was quality and safety; the aim was to identify the managerial activities/time/engagement in quality and safety. The full inclusion/exclusion criteria and screening tool can be accessed in the online Appendices 2-3. Figure 1 presents the numbers of articles included and excluded at each stage of the review process. Four hundred and twenty-three articles remained for full text screening. One reviewer (AP) screened all articles and a second reviewer (AR) reviewed 7% of these. A good agreement inter-rater reliability score was calculated (K=0.615,P<0.001) with 73% agreement. Hand searching and cross-referencing were carried out in case articles were missed by the search strategy or from restriction of databases. One additional article was identified from hand searching, [31] totalling 19 articles included in the systematic review. Review stages based on PRISMA Flow Diagram [32] Records identified through database searching (n = 21,899) Additional records identified through other sources (n = 50) [

Data extraction & methodological quality
The characteristics and summary findings of the 19 included studies are presented in Table   1. This Table is a simplified version of a standardised template that was used to ensure consistency in data extracted from each article. Each study was assessed using a quality appraisal tool developed by , [33] which comprised of two checklists (qualitative and quantitative). Random included articles (32%) were scored by AW for scoring consistency. All articles were scored on up to 24 questions with a score between 0-2, the total percentage scores are presented in Table 1. Some cumulative evidence bias may results from two larger datasets split into more than one study each. [34][35][36][37] Through a narrative synthesis, we aimed to maintain the original meanings, interpretations and raw data offered by the articles. [38]

RESULTS
This results section provides an overview description of the reviewed studies and their key findings. The findings are considered under four main headings: managerial time spent on quality and safety; managerial quality and safety activities; managerial impact on quality and safety; and contextual factors related to managers' quality and safety role. The section ends with a proposed model to summarise the review findings.

Description of the studies
From the 19 included studies, the majority were carried out and set in the US (14 studies), and investigated senior management and/or Boards (13 studies). Of these, 3 focused on senior managers alone (e.g. Chief Nursing Officers), 9 concentrated on Board managers and 1 included a mixture of managerial levels. Only 3 investigated middle managers (e.g. clinical directorate managers) and 3 examined frontline staff (e.g. unit nurse managers). The  11 settings of the study were mostly Trust or hospital-wide; a few articles were set in specific settings or contexts: elderly care, [39] evidence-based medicine, [40] staff productivity, [41] clinical risk management, [42] and hospital acquired infection prevention. [43] Two studies were involved in specific interventions, [44,45] and 7 studies concentrated specifically on QI rather than quality and safety oversight or routine. [34,39,[44][45][46][47][48] There were a mixture of 6 qualitative design (interviews or focus groups); 8 quantitative survey designs; and 5 mixmethods designs. The majority of authors opted for self-reports and all but one study employed a cross-sectional design. [45] The quality assessment scores ranged between 50-100%, with little use of verification procedures to establish credibility of interpretation and lack of reflexivity of qualitative accounts. Some quantitative results presented descriptives alone and failed to report variances.

Managerial time spent on quality and safety
The studies on Board level managers highlight an inadequate prioritisation of quality and patient safety on the Board agenda and subsequent time spent at Board meetings. Not all hospitals consistently have quality on their Board agenda, for example CEOs and chairpersons across 30 organisations reported that approximately a third of all Board meetings had quality on their agenda, [34] and necessary quality items were not consistently and sometimes never addressed. [35] In all studies examining time spent on quality and safety by the Board, less than half of the total time was spent on quality and safety, [31,37,[47][48][49][50] with a majority of Boards spending 25% or less on quality. [31,37,[48][49][50] Findings imply that this may be too low to have a positive influence on quality and safety, as higher quality performance was demonstrated by Boards that spent above 20%/25% on quality. [48,49] Board members recognised that the usual time spent is insufficient. [47] However, few reported financial goals as more important than quality and safety goals, [31] and health system Boards only spent slightly more time on financial issues than quality. [ 12 issues are noted by studies on frontline managers; specifically that they placed less time and importance on QI, [41] identified as the least discussed topic by clinical managers. [51] Managerial quality and safety activities A broad range of quality-related activities were identified to be undertaken by managers.
These are presented by the following three groupings: strategy-centred; data-centred and culture-centred.

Strategy-centred
Board priority-setting and planning strategies aligned with quality and safety goals were identified as Board managerial actions carried out in several studies. High percentages (over 80% in two studies) of Boards had formally established strategic goals for quality with specific targets, and aimed to create a quality plan integral to their broader strategic agenda. [31,36] Contrary findings however suggest that the Board rarely set the agenda for the discussion on quality, [36] did not provide the ideas for their strategies, [31] and were largely uninvolved in strategic planning for QI. [47] In the latter case, the non-clinical Board managers felt that they held "passive" roles in quality decisions. This is important considering evidence that connects the activity of setting the hospital quality agenda with better performance in process of care and mortality. [37] Additionally, Boards that established goals in four areas of quality and publicly disseminated strategic goals and reported quality information were linked to high hospital performance. [34,37,49] Culture-centred Activities aimed at enhancing patient safety/QI culture emerged from several studies across organisational tiers. [43,46,47,52] Board and senior management's activities included encouraging an organisational culture of QI on norms regarding interdepartmental/multi-  13 disciplinary collaboration and advocating QI efforts to clinicians and fellow senior managers, providing powerful messages of safety commitment and influencing the organisation's patient safety mission. [46,52] Managers at differing levels focused on cultivating a culture of clinical excellence and articulating the organisational culture to staff. [43] Factors to motivate/engage middle and senior management in QI included senior management commitment, provision of resources and managerial role accountability. [39,45] Findings revealed connections between senior management and Board priorities and values with hospital performance and on middle management quality-related activities. Ensuring capacity for high quality standards also appears within the remit of management, as physician credentialing was identified as a Board managers' responsibility in more than one study. [37,47] From this review it is unclear to what degree Board involvement in the credentialing process has a significant impact on quality. [37,40] Data-centred Information on quality and safety are continually supplied to the Board. [50] At all levels of management, activities around quality and safety data or information were recognised in 6 studies. [34,37,42,44,46,52] Activities included collecting and collating information, [42] reviewing quality information, [34,37,52] using measures such as incident reports and infection rates to forge changes, [52] using patient satisfaction surveys, [34] taking corrective action based on adverse incidents or trends emphasised at Board meetings, [37] and providing feedback. [42,46] The studies do not specify the changes made based on the data-related activities by senior managers; one study identified that frontline managers predominantly used data from an incident reporting tool to change policy/practice and training/education and communication between care providers. [44] However, overseeing data generally was found to be beneficial, as hospitals that carried out performance monitoring activities had significantly higher scores in process of care and lower mortality rates than hospitals that did not. [37]

Managerial impact on quality and safety outcomes
We have considered the associations found between specific managerial involvement and its affect on quality and safety. Here, we summarise the impact and importance of their general role. Of the articles that looked at either outcomes of management involvement in quality or at its perceived importance, 5 articles suggested that their role was beneficial to quality and safety performance. [31,34,37,39,48] Senior management support and engagement was identified as one of the primary factors associated with good hospital-wide quality outcomes and QI programme success. [34,37,39,48] Conversely, 6 articles suggest that managers' involvement (from the Board, middle and frontline) has little, no or a negative influence on quality and safety. [31,34,37,40,41,43,48] Practices that showed no significant association with quality measures included Board's participation in physician credentialing. [34,37] Another noted that if other champion leaders are present, management leadership was not deemed necessary. [43] Two articles identified a negative or inhibitory effect on evidence-based practices and staff productivity from frontline and middle managers. [40,41]

Contextual factors related to managers' quality and safety role
Most of the articles focussed on issues that influenced the managers' role or their impact, as opposed to discussing the role of the managers. These provide an insight in to the types of conditions in which a manager can best undertake their role to affect quality and safety.
Unfortunately it appears that many of these conditions are not in place.
Two studies found that a Board quality committee is a positive variable in quality performance, but that fewer than 60% had them. [37,49] Similarly, compensation and performance evaluation linked to executive quality performance was identified in 4 articles [34,36,37,48] and associated with better quality performance indicators, [37,48] 15 measures were insufficiently included in CEOs' performance evaluation. [34,36] The use of the right measures to drive QI was raised in relation to Board managerial engagement in quality [34] and to impact on patient care improvement, [50] yet, almost half of this sample did not formally adopt system-wide measures and standards for quality. To aid them in these tasks, evidence indicates the common use of QI measure tools, such as a dashboard or scorecard, [36,48,49] with promising associations between dashboard use and quality outcomes. [37,49] Other factors linked to quality outcomes include management-staff relationship/high interactions between the Board and medical staff when setting quality strategy, [48] and managerial expertise. Although a connection between knowledge with quality outcomes was not found, [37] high performing hospitals have shown higher self-perceived ability to influence care, expertise at the Board and participation in training programs that have a quality component. [49] Disappointingly, there is a low level of CEO knowledge on quality and safety reports, [34] possibly little Boardroom awareness on salient nursing quality issues, [35] and little practice identified to improve quality literacy for the Board. [31,36] There is however promise for new managers through relevant training at induction and by recruitment of those with relevant expertise. [31] The Quality Management IPO Model The input process output (IPO) model is a conceptual framework that helps to structure the review findings in a useful way, please see Figure 2. [53,54]

Perceptions of managers on management
Board practices in quality and safety • Less than half (43%) of Boards reported that they addressed quality and patient safety issues in all meetings • One-third of Boards spend 25% of their time or more on quality and patient safety issues.

•
More than 80% of Boards have formally established strategic goals for quality with specific targets, but a majority of Board chairs indicate that their Boards did not provide the ideas for strategic direction or initiatives. • Board chairs reported a low participation in education on quality and safety: 43% reported that all the Board members participated, 19% stated that more than half participated and 23% said it was less than a quarter of the Board • Most Board chairs (87%) reported Board member induction training on responsibilities for quality and safety, although almost a third (30%) reported few or no opportunities for education on this, 42% reported some opportunities and 28% reported many).

•
Approximately half (57%) of the Board chairs acknowledged recruitment of individuals that have knowledge, skills and experience in quality and patient safety onto the Board. A Board skills matrix included quality and safety as one of the competency areas. 16 Perceptions of managers on their engagement in a QI programme Five elements deemed essential to middle manager engagement: • (1) Senior management commitment and leadership ( e.g. senior management provides strategic direction for QI plan) • (2) Provision of resources and opportunities for QI education and information dissemination (e.g. basic QI skills provided to all staff) • (3) Senior and middle manager role accountability (e.g.: senior managers and middle managers agree QI roles and expectations) • (4) Middle manager involvement in QI planning (e.g. senior and middle managers plan together) • (5) Middle managers own and operate QI program (e.g. ongoing review and evaluation of the progress of the QI program by the middle and senior managers) 14 Senior management support reported as the primary enabling factor in the implementation of such programmes (96.6%), along with a lack of support as the primary reason for not implementing the programme (65.0%) • The interviews supported that having an administrative champion was considered essential to their program's success 19

Perceptions of managers on managers'
quality and safety practices

•
The majority of managers (91%) who received collated incident information used it to feed back to their own staff. 60% always fed back to staff, 28% sometimes did, 2% never did.

•
Of the Trusts that had written guidance on types of clinical incident to report, 80% of managers had general guidance and fewer (20%) had written speciality specific guidance. • 76% of managers reported information collation of clinical incidents. Of these, 59% were involved in data collection themselves.

DISCUSSION
Our review examined the role of managers in maintaining and promoting safe, quality care.
The existing studies detail the time spent, activities and engagement of hospital managers and Boards, and suggest that these can positively influence quality and safety performance.
They further reveal that such involvement is often absent, as are certain conditions that may help them in their work.
Evidence promotes hospitals to have a Board quality committee, with a specific item on quality at the Board meeting, a quality performance measurement report and a dashboard with national quality and safety benchmarks along with standardised quality and safety measures. Outside of the Boardroom, the implications are for senior managers to build a good infrastructure for staff-manager interactions on quality strategies and attach compensation and performance evaluation to quality and safety achievements. For QI programmes, managers should keep in mind its consistency with the hospital's mission and provide commitment, resources, education, and role accountability. Literature elsewhere supports much of these findings, such as the use of quality measurement tools [21,55] better quality-associated compensation, a separate quality committee, [16,56] and has also emphasised poor manager-clinician relationships as damaging to patients and QI. [57,58] Some of the variables that were shown to be associated with good quality performance, such as having a Board committee, compensation/performance and adoption of system-wide • There is a dearth of empirical evidence on hospital managerial work and its influence on quality of care.
• There is some evidence that Boards'/managers' time spent, engagement and work can influence quality and safety clinical outcomes, processes and performance.
• Some variables associated with good quality performance were lacking in study hospitals.
• Many Board managers do not spend sufficient time on quality and safety and need to develop their knowledge on quality and safety.
• There is a greater focus on the contextual issues surrounding managers' roles than on examining managerial activities.
• Research is required to examine middle and frontline managers, to take into consideration non-managers' perceptions, and to assess senior managers' time and tasks outside of the Boardroom • We present a model to summarise the evidence-based promotion of conditions and activities for managers to best affect quality performance.  30 We would like to thank Miss Dina Grishin for helping to review the abstracts and Miss Ana Wheelock for helping to assess the quality of the articles.

COMPETING INTERESTS
There are no competing interests.

2
Acute Hospitals/ or hospital care/ or (hospital$ or secondary care or acute care or health care organi*ation$1 or healthcare organi*ation$1 or infirmar$).ti,ab.

3
exp managers/ or "middle and lower management"/ or senior managers/ or top management/ or (Manager$1 or Matron$1 or CEO$1 or executive$1 or director$3 or board$1 or middle management or senior management or lower management or frontline management or leader$4 or president$1 or head of department$1 or department head$1 or head of nursing or administrator$1 or healthcare administration or (chief adj4 officer$1) or (chief adj4 nurs$) or (chief adj4 operation$1) or (chief adj4 service$1) or chief of staff).ti,ab.  Key Terms Quality: Health Care Quality or length of stay or mortality or hospital readmission or exp evidence based medicine or evidence based practice or exp outcome assessment or quality control or medical audit or patient satisfaction or patient centred care or length of stay or mortality or customer satisfaction or patient readmission or evidence based medicine or or waiting times or patient experience or complaints or target(s) or clinical excellence or service excellence or quality or patient safety or medical errors or adverse events Not: Cost C) HOSPITALS

Protocol and registration
5 Indicate if a review protocol exists, if and where it can be accessed (e.g., Web address), and, if available, provide registration information including registration number.
All protocol info in appendices Eligibility criteria 6 Specify study characteristics (e.g., PICOS, length of follow-up) and report characteristics (e.g., years considered, language, publication status) used as criteria for eligibility, giving rationale.

7
Information sources 7 Describe all information sources (e.g., databases with dates of coverage, contact with study authors to identify additional studies) in the search and date last searched.

6
Search 8 Present full electronic search strategy for at least one database, including any limits used, such that it could be repeated.
in appendices Study selection 9 State the process for selecting studies (i.e., screening, eligibility, included in systematic review, and, if applicable, included in the meta-analysis).

7
Data collection process 10 Describe method of data extraction from reports (e.g., piloted forms, independently, in duplicate) and any processes for obtaining and confirming data from investigators.

10
Data items 11 List and define all variables for which data were sought (e.g., PICOS, funding sources) and any assumptions and simplifications made.

6
Risk of bias in individual studies 12 Describe methods used for assessing risk of bias of individual studies (including specification of whether this was done at the study or outcome level), and how this information is to be used in any data synthesis.

DISCUSSION
Summary of evidence 24 Summarize the main findings including the strength of evidence for each main outcome; consider their relevance to key groups (e.g., healthcare providers, users, and policy makers).

24
Limitations 25 Discuss limitations at study and outcome level (e.g., risk of bias), and at review-level (e.g., incomplete retrieval of identified research, reporting bias).     Conclusion: There is some evidence that managers' time spent and work can influence quality and safety clinical outcomes, processes and performance. However, there is a dearth of empirical studies, further weakened by a lack of objective outcome measures and little examination of actual actions undertaken. We present a model to summarise the conditions and activities that affect quality performance.  • The review reveals conditions and actions conducive to good quality management and offers a model to transparently present these to managers considering their own part in quality and safety.
• The search for this review has screened a vast amount of the literature (over 15,000 articles) across a number of databases.
• The small number of included studies and their varied study aims, design and population samples make generalisations difficult. With more literature on this topic, distinctions could be made between job positions.
• The quality assessment scores are subjective and may not take into consideration factors beyond the quality assessment scale used.

INTRODUCTION
Managers in healthcare have a legal and moral obligation to ensure a high quality of patient care and to strive to improve care. These managers are in a prime position to mandate policy, systems, procedures and organisational climates. Accordingly, many have argued that it is evident that healthcare managers possess an important and obvious role in quality of care and patient safety and that it is one of the highest priorities of healthcare managers. [1][2][3] In line with this, there have been calls for Boards to take responsibility for quality and safety outcomes. [4,5] One article warned hospital leaders of the dangers of following in the path of bankers falling into recession, constrained by their lack of risk awareness and reluctance to take responsibility. [6] To add to the momentum are some high profile publicity of hospital management failures affecting quality and safety, eliciting strong instruction for managerial leadership for quality at the national level in some countries. [7,8] Beyond healthcare, there is clear evidence of managerial impact on workplace safety. [9][10][11][12] Within the literature on healthcare, there are non-empirical articles providing propositions and descriptions on managerial attitudes and efforts to improve safety and quality. This literature, made up of opinion articles, editorials and single participant experiences, present an array of insightful suggestions and recommendations for actions that hospital managers should take to improve the quality of patient care delivery in their organisation. [13][14][15][16][17] However, researchers have indicated that there is a limited evidence-base on this topic. [18][19][20][21] Others highlight the literature focus on the difficulties of the managers' role and the negative results of poor leadership on quality improvement (QI) rather than considering actions that managers presently undertake on quality and safety. [22,23]   How much time is spent by hospital managers on quality and safety and its improvement?
What are the managerial activities that relate to quality and safety and its improvement?
How are managers engaged in quality and safety and its improvement?
What impact do managers have on quality and safety and its improvement?
How do contextual factors influence the managers' role and impact on quality and safety and its improvement?   following dimensions: safe, effective, patient-centred, timely, efficient and equitable. [4] They define patient safety simply as "the prevention of harm to patients", [24] and AHRQ define it is "freedom from accidental or preventable injuries produced by medical care." [25] Literature was searched for all key terms associated with quality and patient safety to produce an allencompassing approach. A manager was defined as an employee that has subordinates, oversees staff, is responsible for staff recruitment and training, and holds budgetary accountabilities. Therefore, all levels of managers including Boards of managers were included in this review with the exception of clinical frontline employees, e.g. doctors or nurses, who may have taken on further managerial responsibilities alongside their work but do not have a primary official role as a manager. Those that have specifically taken on a role for quality of care, e.g. the modern matron, were also excluded. Distinction between senior, middle and frontline management were as follows: senior management hold Trust-wide responsibilities; [26] middle managers are in the middle of the organisational hierarchy chart and have one or more managers reporting to them; [27] frontline managers are defined as managers at the first level of the organisational hierarchy chart who have frontline employees reporting to them. Board managers include all members of the Board. Although, there are overlaps between senior managers and Boards (for example Chief Executive Officers (CEOs) may sit on hospital Boards), we aim to present senior and Board level managers separately due to the differences in their responsibilities and position. Only managers that would manage within or govern hospitals were included, with the exclusion of settings that solely served mental health or that comprised solely of non-acute care community services (in order to keep the sample more homogenous). The definition of 'Role' focused on actual engagement, time spent and activities that do or did occur rather than those recommended that should or could occur.

Search Strategy
Literature was reviewed between 01 Jan 1983 and 01 Nov 2010. Eligible articles were those that described or tested managerial roles pertaining to quality and safety in the hospital setting. Part of the search strategy was based on guidance by . [28] EMBASE, MEDLINE, Health Management Information Consortium (HMIC) and PSYCHINFO databases were searched. The search strategy involved three facets (i.Management; ii.Quality; iii.Hospital Setting) and five steps. A facet (i.e. a conceptual grouping of related search terms) for role was not included in the search strategy, as it would have significantly reduced the sensitivity of the search.
Multiple iterations and combinations of all search terms were tested to achieve the best level of specificity and sensitivity. In addition to the key terms, Medical Subject Headings (MeSH®) terms were used, which were 'exploded' to include all MeSH subheadings. All databases required slightly different MeSH terms (named Emtree in Embase), therefore four variations of the search strategies were used (see online Appendix 1 for the search strategies). Additional limits placed on the search strategy restricted study subjects to human and the language to English. The search strategy identified 15,447 articles after duplicates had been removed.

Screening
Three reviewers (AP, AR and DG) independently screened the titles and abstracts of the articles for studies that fit the inclusion criteria. One reviewer (AP) screened all 15,447 articles, while two additional reviewers screened 30% of the total sample retrieved from the search strategy: AR screened 20% and DG screened 10%. On testing inter-rater reliability, Cohen's kappa correlations showed low agreement between AR and AP (K=0.157, P<0.01) and between DG and AP (K=0.137, P<0.00). [29] However, there was a high percentage of agreement between raters (95% and 89% respectively), which reveals a good inter-rater  [30,31] Discrepancies were resolved by discussion and consensus. The main inclusion criteria were that: the setting was hospitals; the population sample reported on were managers; the context was quality and safety; the aim was to identify the managerial activities/time/engagement in quality and safety. The full inclusion/exclusion criteria and screening tool can be accessed in the online Appendices 2-3. Figure 1 presents the numbers of articles included and excluded at each stage of the review process.
Four hundred and twenty-three articles remained for full text screening. One reviewer (AP) screened all articles and a second reviewer (AR) reviewed 7% of these. A moderate agreement inter-rater reliability score was calculated (K=0.615,P<0.001) with 73% agreement. The primary reoccurring difference in agreement was regarding whether the article pertained to quality of care, owing to the broad nature of the definition. Each article was discussed individually until a consensus was reached on whether to include or exclude.
Hand searching and cross-referencing were carried out in case articles were missed by the search strategy or from restriction of databases. One additional article was identified from hand searching, [32] totalling 19 articles included in the systematic review.

Data extraction & methodological quality
The characteristics and summary findings of the 19 included studies are presented in Table   1. This Table is a simplified version of a standardised template that was used to ensure consistency in data extracted from each article. Each study was assessed using a quality appraisal tool developed by , [34] which comprised of two checklists (qualitative and quantitative). Random included articles (32%) were scored by AW for scoring consistency. All articles were scored on up to 24 questions with a score between 0-2; Box 1 shows an example definition of what constitutes 'Yes' (2), 'Partial' (1) and 'No' (0) rating criteria. The total percentage scores for each study are presented in Table 1. All studies were included regardless of their quality scores. Some cumulative evidence bias may results from two larger datasets split into more than one study each. [35][36][37][38] Through a narrative synthesis, we aimed to maintain the original meanings, interpretations and raw data offered by the articles. [39] Rating Criteria to verify whether question or objective is sufficiently described

Yes
Is easily identified in the introductory section (or first paragraph of methods section).
Specifies (where applicable, depending on study design) all of the following: purpose, subjects/target population, and the specific intervention(s) /association(s)/descriptive parameter(s) under investigation. A study purpose that only becomes apparent after studying other parts of the paper is not considered sufficiently described.
Partial Vaguely/incompletely reported (e.g. "describe the effect of" or "examine the role of" or "assess opinion on many issues" or "explore the general attitudes"...); or some information has to be gathered from parts of the paper other than the introduction/background/objective section.

No
Question or objective is not reported, or is incomprehensible.

N/A Should not be checked for this question
Box 1 Example of a rating criteria for Kmet's quality assessment [34]

RESULTS
This results section provides an overview description of the reviewed studies and their key findings. The findings are considered under four main headings: managerial time spent on quality and safety; managerial quality and safety activities; managerial impact on quality and safety; and contextual factors related to managers' quality and safety role. The section ends with a proposed model to summarise the review findings.

Description of the studies
From the 19 included studies, the majority were carried out and set in the US (14 studies), and investigated senior management and/or Boards (13 studies). Of these, 3 focused on senior managers alone (e.g. Chief Nursing Officers), 9 concentrated on Board managers and 1 included a mixture of managerial levels. Only 3 investigated middle managers and 3 examined frontline staff (e.g. clinical directorate managers and unit nurse managers). The settings of the study were mostly Trust or hospital-wide; a few articles were set in specific settings or contexts: elderly care, [40] evidence-based medicine, [41] staff productivity, [42] clinical risk management, [43] and hospital acquired infection prevention. [44] Two studies involved specific interventions, [45,46] and 7 studies concentrated specifically on QI rather than quality and safety oversight or routine. [35,40,[45][46][47][48][49] There were a mixture of 6 qualitative design (interviews or focus groups); 8 quantitative survey designs; and 5 mixmethods designs. All but one study employed a cross-sectional design [46]. [46] The primary outcome measure used in most studies was perceptions of managerial quality and safety practices. All reported participant perceptions and a majority presented self-reports, that is, either a mixture of self and peer reports, or self-reports alone. [41,43,45,46] Several studies asked participants about their own and/or other managers' involvement in regards to their specific quality improvement intervention or quality/safety issue. [40,41,[44][45][46][47] With some variations, the most common research design was to interview or survey senior manager/Board members (particularly Board chairs, presidents and CEOs) perceptions on questions were asked of participants, or no mention of consent and confidentiality assurances. In 7 studies there was no or vague qualitative data analysis description, including omitting the type of qualitative analysis used. Six of the studies showed no or poor use of verification procedures to establish credibility and 9 reported no or poor reflexivity.
Positively, all study designs were evident, the context of studies were clear and the authors showed a connection to a wider body of knowledge.
Similarly to the qualitative studies, 7 quantitative studies did not fully describe, justify or use appropriate analysis methods. However, compared with the qualitative studies, the quantitative studies suffered more from sampling issues. Three studies had particularly small samples (e.g. n=35) and one had an especially low response rate of 15%. Subject characteristics were insufficiently described in 5 studies; in one case the authors did not state the number of hospitals included in data analysis. Several studies had obtained ordinal data but only presented percentages, and only one study appropriately controlled for confounding variables. Across all articles, all but 3 studies reported clear objectives and asserted conclusions clearly supported by the data.

Managerial time spent on quality and safety
The studies on Board level managers highlight an inadequate prioritisation of quality and patient safety on the Board agenda and subsequent time spent at Board meetings. Not all hospitals consistently have quality on their Board agenda, for example CEOs and chairpersons across 30 organisations reported that approximately a third of all Board meetings had quality on their agenda, [35] and necessary quality items were not consistently and sometimes never addressed. [36] In all studies examining time spent on quality and safety by the Board, less than half of the total time was spent on quality and safety, [32,37,38,[48][49][50][51] with a majority of Boards spending 25% or less on quality. [32,38,45,[49][50][51] Findings imply that this may be too low to have a positive influence on quality and safety, as higher quality performance was demonstrated by Boards that spent above 20%/25% on quality. [49,50] Board members recognised that the usual time spent is insufficient. [48] However, few reported financial goals as more important than quality and safety goals, [32] and health system Boards only spent slightly more time on financial issues than quality. [51] Similar issues are noted by studies on frontline managers; specifically that they placed less time and importance on QI, [42] identified as the least discussed topic by clinical managers. [52] Managerial quality and safety activities A broad range of quality-related activities were identified to be undertaken by managers.
These are presented by the following three groupings: strategy-centred; data-centred and culture-centred.

Strategy-centred
Board priority-setting and planning strategies aligned with quality and safety goals were identified as Board managerial actions carried out in several studies. High percentages (over 80% in two studies) of Boards had formally established strategic goals for quality with specific targets, and aimed to create a quality plan integral to their broader strategic agenda. [32,37] Contrary findings however suggest that the Board rarely set the agenda for the discussion on quality, [37] did not provide the ideas for their strategies, [32] and were largely uninvolved in strategic planning for QI. [48] In the latter case, the non-clinical Board managers felt that they held "passive" roles in quality decisions. This is important considering evidence that connects the activity of setting the hospital quality agenda with better performance in process of care and mortality. [38] Additionally, Boards that established goals in four areas of quality and publicly disseminated strategic goals and reported quality information were linked to high hospital performance. [35,38,50] Culture-centred Activities aimed at enhancing patient safety/QI culture emerged from several studies across organisational tiers. [44,47,48,53] Board and senior management's activities included encouraging an organisational culture of QI on norms regarding interdepartmental/multidisciplinary collaboration and advocating QI efforts to clinicians and fellow senior managers, providing powerful messages of safety commitment and influencing the organisation's patient safety mission. [47,53] Managers at differing levels focused on cultivating a culture of clinical excellence and articulating the organisational culture to staff. [44] Factors to motivate/engage middle and senior management in QI included senior management commitment, provision of resources and managerial role accountability. [40,46] Findings revealed connections between senior management and Board priorities and values with hospital performance and on middle management quality-related activities. Ensuring capacity for high quality standards also appears within the remit of management, as  14 physician credentialing was identified as a Board managers' responsibility in more than one study. [38,48] From this review it is unclear to what degree Board involvement in the credentialing process has a significant impact on quality. [38,41] Data-centred Information on quality and safety are continually supplied to the Board. [51] At all levels of management, activities around quality and safety data or information were recognised in 6 studies. [35,38,43,45,47,53] Activities included collecting and collating information, [43] reviewing quality information, [35,38,53] using measures such as incident reports and infection rates to forge changes, [53] using patient satisfaction surveys, [35] taking corrective action based on adverse incidents or trends emphasised at Board meetings, [38] and providing feedback. [43,47] The studies do not specify the changes made based on the data-related activities by senior managers; one study identified that frontline managers predominantly used data from an incident reporting tool to change policy/practice and training/education and communication between care providers. [45] However, overseeing data generally was found to be beneficial, as hospitals that carried out performance monitoring activities had significantly higher scores in process of care and lower mortality rates than hospitals that did not. [38] Managerial impact on quality and safety outcomes We have considered the associations found between specific managerial involvement and its affect on quality and safety. Here, we summarise the impact and importance of their general role. Of the articles that looked at either outcomes of management involvement in quality or at its perceived importance, 6 articles suggested that their role was beneficial to quality and safety performance. [32,35,38,40,49,53] Senior management support and engagement was identified as one of the primary factors associated with good hospital-wide quality outcomes and QI programme success. [35,38,40,49] Conversely, 6 articles suggest  15 that managers' involvement (from the Board, middle and frontline) has little, no or a negative influence on quality and safety. [35,38,41,42,44,49] Practices that showed no significant association with quality measures included Board's participation in physician credentialing. [35,38] Another noted that if other champion leaders are present, management leadership was not deemed necessary. [44] Two articles identified a negative or inhibitory effect on evidence-based practices and staff productivity from frontline and middle managers. [41,42]

Contextual factors related to managers' quality and safety role
Most of the articles focussed on issues that influenced the managers' role or their impact, as opposed to discussing the role of the managers. These provide an insight in to the types of conditions in which a manager can best undertake their role to affect quality and safety.
Unfortunately it appears that many of these conditions are not in place.
Two studies found that a Board quality committee is a positive variable in quality performance, but that fewer than 60% had them. [38,50] Similarly, compensation and performance evaluation linked to executive quality performance was identified in 4 articles [35,37,38,49] and associated with better quality performance indicators, [38,49] but quality measures were insufficiently included in CEOs' performance evaluation. [35,37] The use of the right measures to drive QI was raised in relation to Board managerial engagement in quality [35] and to impact on patient care improvement, [51] yet, almost half of this sample did not formally adopt system-wide measures and standards for quality. To aid them in these tasks, evidence indicates the common use of QI measure tools, such as a dashboard or scorecard, [37,49,50] with promising associations between dashboard use and quality outcomes. [38,50] 16 Other factors linked to quality outcomes include management-staff relationship/high interactions between the Board and medical staff when setting quality strategy, [49] and managerial expertise. Although a connection between knowledge with quality outcomes was not found, [38] high performing hospitals have shown higher self-perceived ability to influence care, expertise at the Board and participation in training programs that have a quality component. [50] Disappointingly, there is a low level of CEO knowledge on quality and safety reports, [35] possibly little Boardroom awareness on salient nursing quality issues, [36] and little practice identified to improve quality literacy for the Board. [32,37] There is however promise for new managers through relevant training at induction and by recruitment of those with relevant expertise. [32] The Quality Management IPO Model The input process output (IPO) model is a conceptual framework that helps to structure the review findings in a useful way, please see Figure   Self-reported perceptions of managers on managers' quality and safety practices

•
The majority of managers (91%) who received collated incident information used it to feed back to their own staff. 60% always fed back to staff, 28% sometimes did, 2% never did.

•
Of the Trusts that had written guidance on types of clinical incident to report, 80% of managers had general guidance and fewer (20%) had written speciality specific guidance.

DISCUSSION
Our review examined the role of managers in maintaining and promoting safe, quality care.
The existing studies detail the time spent, activities and engagement of hospital managers and Boards, and suggest that these can positively influence quality and safety performance.
They further reveal that such involvement is often absent, as are certain conditions that may help them in their work.
Evidence from the review promotes hospitals to have a Board quality committee, with a specific item on quality at the Board meeting, a quality performance measurement report and a dashboard with national quality and safety benchmarks along with standardised quality and safety measures. Outside of the Boardroom, the implications are for senior managers to build a good infrastructure for staff-manager interactions on quality strategies and attach compensation and performance evaluation to quality and safety achievements. For QI programmes, managers should keep in mind its consistency with the hospital's mission and provide commitment, resources, education, and role accountability. Literature elsewhere supports much of these findings, such as the use of quality measurement tools [21,56] better quality-associated compensation, a separate quality committee, [16,57] and has also emphasised poor manager-clinician relationships as damaging to patients and QI. [58,59] Some of the variables that were shown to be associated with good quality performance, such as having a Board committee, compensation/performance and adoption of system-wide have on quality and safety (rather than examination of their role). These studies have shown senior managerial leadership to be associated with a higher degree of QI implementation, [65] promotion of clinical involvement, [66,67] safety climate attitudes, [68] and increased Board leadership for quality. Research on this area is particularly required to examine middle and frontline managers, to take into consideration non-managers' perceptions, and to assess senior managers' time

Review limitations
There which ultimately reduces the validity of the conclusions drawn from their findings. As most of the study findings relied on self-reports, social desirability may have resulted in exaggerated processes and inflated outputs. Although, encouragingly, one of the included studies found that managers that perceived their Boards to be effective in quality oversight were from hospitals that had higher processes-of-care scores and lower risk adjusted mortality. The quality assessment scores should be viewed with caution; such scores are subjective and may not take into consideration factors beyond the quality assessment scale used. Due to the enormity of this review, the publication of this article is some time after the search run date. As there is little evidence published on this topic, we consider this not to greatly impact • There is a dearth of empirical evidence on hospital managerial work and its influence on quality of care.
• There is some evidence that Boards'/managers' time spent, engagement and work can influence quality and safety clinical outcomes, processes and performance.
• Some variables associated with good quality performance were lacking in study hospitals.
• Many Board managers do not spend sufficient time on quality and safety.
• There is a greater focus on the contextual issues surrounding managers' roles than on examining managerial activities.
• Research is required to examine middle and frontline managers, to take into consideration non-managers' perceptions, and to assess senior managers' time and tasks outside of the Boardroom. More robust methodologies with objective outcome measures would strengthen the evidence.
• We present a model to summarise the evidence-based promotion of conditions and activities for managers to best affect quality performance.

ACKNOWLEDGEMENTS
We would like to thank Miss Dina Grishin for helping to review the abstracts and Miss Ana Wheelock for helping to assess the quality of the articles.

COMPETING INTERESTS
There are no competing interests.

FUNDING
This work was supported by funding from the Health Foundation and the National Institute for Health Research.    Conclusion: There is some evidence that managers' time spent and work can influence quality and safety clinical outcomes, processes and performance. However, there is a dearth of empirical studies on their work and its influence, further weakened by a lack of objective  1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58 1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58   In line with this, there have been calls for Boards to take responsibility for quality and safety outcomes. [4,5] One article warned hospital leaders of the dangers of following in the path of bankers falling into recession, constrained by their lack of risk awareness and reluctance to take responsibility. [6] To add to the momentum are some high profile publicity of hospital management failures affecting quality and safety, eliciting strong instruction for managerial leadership for quality at the national level in some countries. [7,8] Beyond healthcare, there is clear evidence of managerial impact on workplace safety. [9][10][11][12] Within the literature on healthcare, there are non-empirical articles providing propositions and descriptions on managerial attitudes and efforts to improve safety and quality. This literature, made up of opinion articles, editorials and single participant experiences, present an array of insightful suggestions and recommendations for actions that hospital managers should take to improve the quality of patient care delivery in their organisation. [13][14][15][16][17] However, researchers have indicated that there is a limited evidence-base on this topic. [18][19][20][21] Others highlight the literature focus on the difficulties of the managers' role and the negative results of poor leadership on quality improvement (QI) rather than considering actions that managers presently undertake on quality and safety. [22,23]   How much time is spent by hospital managers on quality and safety and its improvement?

Figure Legends
What are the managerial activities that relate to quality and safety and its improvement?
How are managers engaged in quality and safety and its improvement?
What impact do managers have on quality and safety and its improvement?
How do contextual factors influence the managers' role and impact on quality and safety and its improvement?   1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58   following dimensions: safe, effective, patient-centred, timely, efficient and equitable. [4] They define patient safety simply as "the prevention of harm to patients", [24] and AHRQ define it is "freedom from accidental or preventable injuries produced by medical care." [25] and lLiterature was searched for all key terms associated with quality and patient safety to produce an all encompassingall-encompassing approach. A manager was defined as any employee that manages has subordinates, oversees staff and, is likely to holdis managerial responsibilitiesle such asfor budget responsibilities and staff recruitment and training, and holds budgetary accountabilities. Therefore, all levels of managers including Boards of managers were included in this review with the exception of clinical frontline employees, e.g.

Concepts and definitions
doctors or nurses, who may have taken on further managerial responsibilities alongside their work but do not have a primary official role as a manager. Those that have specifically taken on a role for quality of care, e.g. the modern matron, were also excluded. Distinction between senior, middle and frontline management were as follows: senior management hold Trust-wide responsibilities; [26] middle managers are in the middle of the organisational hierarchy chart and have one or more managers reporting to them; [27] frontline managers are defined as managers at the first level of the organisational hierarchy chart who have frontline employees reporting to them. Board managers include all members of the Board.
Although, there are overlaps between senior managers and Boards (for example Chief Executive Officers (CEOs) may sit on hospital Boards), we aim to present senior and Board level managers separately due to the differences in their responsibilities and position. Only managers that would manage within or govern hospitals were included, with the exclusion of settings that solely served mental health or that comprised solely of non-acute care community services (in order to keep the sample more homogenous). The definition of 'Role' focused on actual engagement, time spent and activities that do or did occur rather than those recommended that should or could occur..  1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58

Screening
Three reviewers (AP, AR and DG) independently screened the titles and abstracts of the articles for studies that fit the inclusion criteria. One reviewer (AP) screened all 15,447 articles, while two additional reviewers screened 30% of the total sample retrieved from the search strategy: AR screened 20% and DG screened 10%. On testing inter-rater reliability, Cohen's kappa correlations showed low agreement between AR and AP (K=0.157, P<0.01) and between DG and AP (K=0.137, P<0.00). [29] However, there was a high percentage of agreement between raters (95% and 89% respectively), which reveals a good inter-rater   1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58 [30,31] Discrepancies were resolved by discussion and consensus. The main inclusion criteria were that: the setting was hospitals; the population sample reported on were managers; the context was quality and safety; the aim was to identify the managerial activities/time/engagement in quality and safety. The full inclusion/exclusion criteria and screening tool can be accessed in the online Appendices 2-3. Figure 1 presents the numbers of articles included and excluded at each stage of the review process.
Four hundred and twenty-three articles remained for full text screening. One reviewer (AP) screened all articles and a second reviewer (AR) reviewed 7% of these. A good moderate agreement inter-rater reliability score was calculated (K=0.615,P<0.001) with 73% agreement. The primary reoccurring difference in agreement was regarding whether the article pertained to quality of care, dueowing to the broad nature of the definition. Each article was discussed individually until a consensus was reached on whether to include or exclude. Hand searching and cross-referencing were carried out in case articles were missed by the search strategy or from restriction of databases. One additional article was identified from hand searching, [32] totalling 19 articles included in the systematic review.

Data extraction & methodological quality
The characteristics and summary findings of the 19 included studies are presented in Table   1. This Table is a simplified version of a standardised template that was used to ensure consistency in data extracted from each article. Each study was assessed using a quality appraisal tool developed by , [34] which comprised of two checklists (qualitative and quantitative). Random included articles (32%) were scored by AW for  1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58 Table 1. Box 1 shows an example definition of what constitutes 'Yes' (2), 'Partial' (1) and 'No' (0) rating criteria. The total percentage scores for each study are presented in Table 1. All studies were included regardless of their quality scores. Some cumulative evidence bias may results from two larger datasets split into more than one study each. [35][36][37][38] Through a narrative synthesis, we aimed to maintain the original meanings, interpretations and raw data offered by the articles. [39] Rating Criteria to verify whether question or objective is sufficiently described

Yes
Is easily identified in the introductory section (or first paragraph of methods section).
Specifies (where applicable, depending on study design) all of the following: purpose, subjects/target population, and the specific intervention(s) /association(s)/descriptive parameter(s) under investigation. A study purpose that only becomes apparent after studying other parts of the paper is not considered sufficiently described.
Partial Vaguely/incompletely reported (e.g. "describe the effect of" or "examine the role of" or "assess opinion on many issues" or "explore the general attitudes"...); or some information has to be gathered from parts of the paper other than the introduction/background/objective section.

No
Question or objective is not reported, or is incomprehensible.
These are presented by the following three groupings: strategy-centred; data-centred and culture-centred.  [38,48] From this review it is unclear to what degree Board involvement in the credentialing process has a significant impact on quality. [38,41] Data-centred Information on quality and safety are continually supplied to the Board. [51] At all levels of management, activities around quality and safety data or information were recognised in 6 studies. [35,38,43,45,47,53] Activities included collecting and collating information, [43] reviewing quality information, [35,38,53] using measures such as incident reports and infection rates to forge changes, [53] using patient satisfaction surveys, [35] taking corrective action based on adverse incidents or trends emphasised at Board meetings, [38] and providing feedback. [43,47] The studies do not specify the changes made based on the data-related activities by senior managers; one study identified that frontline managers predominantly used data from an incident reporting tool to change policy/practice and training/education and communication between care providers. [45] However, overseeing data generally was found to be beneficial, as hospitals that carried out performance monitoring activities had significantly higher scores in process of care and lower mortality rates than hospitals that did not. [38] Managerial impact on quality and safety outcomes We have considered the associations found between specific managerial involvement and its affect on quality and safety. Here, we summarise the impact and importance of their general role. Of the articles that looked at either outcomes of management involvement in quality or at its perceived importance, 5 6 articles suggested that their role was beneficial to quality and safety performance. [32,35,38,40,49,53] Senior management support and engagement was identified as one of the primary factors associated with good hospital-wide quality outcomes and QI programme success. [35,38,40,49] Conversely, 6 articles suggest

Contextual factors related to managers' quality and safety role
Most of the articles focussed on issues that influenced the managers' role or their impact, as opposed to discussing the role of the managers. These provide an insight in to the types of conditions in which a manager can best undertake their role to affect quality and safety.
Unfortunately it appears that many of these conditions are not in place.
Two studies found that a Board quality committee is a positive variable in quality performance, but that fewer than 60% had them. [38,50] Similarly, compensation and performance evaluation linked to executive quality performance was identified in 4 articles [35,37,38,49] and associated with better quality performance indicators, [38,49] but quality measures were insufficiently included in CEOs' performance evaluation. [35,37] The use of the right measures to drive QI was raised in relation to Board managerial engagement in quality [35] and to impact on patient care improvement, [51] yet, almost half of this sample did not formally adopt system-wide measures and standards for quality. To aid them in these tasks, evidence indicates the common use of QI measure tools, such as a dashboard or scorecard, [37,49,50] with promising associations between dashboard use and quality outcomes. [38,50]

The Quality Management IPO Model
The input process output (IPO) model is a conceptual framework that helps to structure the review findings in a useful way, please see Figure 2.  The majority of managers (91%) who received collated incident information used it to feed back to their own staff. 60% always fed back to staff, 28% sometimes did, 2% never did.

•
Of the Trusts that had written guidance on types of clinical incident to report, 80% of managers had general guidance and fewer (20%) had written speciality specific guidance. • 76% of managers reported information collation of clinical incidents. Of these, 59% were involved in data collection themselves.    The existing studies detail the time spent, activities and engagement of hospital managers and Boards, and suggest that these can positively influence quality and safety performance.

NA 13/22 (59%) Activities
They further reveal that such involvement is often absent, as are certain conditions that may help them in their work.
Evidence from the review promotes hospitals to have a Board quality committee, with a specific item on quality at the Board meeting, a quality performance measurement report and a dashboard with national quality and safety benchmarks along with standardised quality and safety measures. Outside of the Boardroom, the implications are for senior managers to build a good infrastructure for staff-manager interactions on quality strategies and attach compensation and performance evaluation to quality and safety achievements. For QI programmes, managers should keep in mind its consistency with the hospital's mission and provide commitment, resources, education, and role accountability. Literature elsewhere supports much of these findings, such as the use of quality measurement tools [21,56] better quality-associated compensation, a separate quality committee, [16,57] and has also emphasised poor manager-clinician relationships as damaging to patients and QI. [58,59] Some of the variables that were shown to be associated with good quality performance, such as having a Board committee, compensation/performance and adoption of system-wide measures, were lacking within the study hospitals. There are also indications of the need to develop Board and senior managerial knowledge and training on quality and safety. have on quality and safety (rather than examination of their role). These studies have shown senior managerial leadership to be associated with a higher degree of QI implementation, [65] promotion of clinical involvement, [66,67] safety climate attitudes, [68] and increased Board leadership for quality.      • There is a dearth of empirical evidence on hospital managerial work and its influence on quality of care.
• • There is some evidence that Boards'/managers' time spent, engagement and work can influence quality and safety clinical outcomes, processes and performance.
• Some variables associated with good quality performance were lacking in study hospitals.
• Many Board managers do not spend sufficient time on quality and safety and need to develop their knowledge on quality and safety.
• There is a greater focus on the contextual issues surrounding managers' roles than on examining managerial activities.

ACKNOWLEDGEMENTS
We would like to thank Miss Dina Grishin for helping to review the abstracts and Miss Ana Wheelock for helping to assess the quality of the articles.

DISCUSSION
Summary of evidence 24 Summarize the main findings including the strength of evidence for each main outcome; consider their relevance to key groups (e.g., healthcare providers, users, and policy makers).

24
Limitations 25 Discuss limitations at study and outcome level (e.g., risk of bias), and at review-level (e.g., incomplete retrieval of identified research, reporting bias).     1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58 1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58 -3] In line with this, there have been calls for Boards to take responsibility for quality and safety outcomes. [4,5] One article warned hospital leaders of the dangers of following in the path of bankers falling into recession, constrained by their lack of risk awareness and reluctance to take responsibility. [6] To add to the momentum are some high profile publicity of hospital management failures affecting quality and safety, eliciting strong instruction for managerial leadership for quality at the national level in some countries. [7,8] Beyond healthcare, there is clear evidence of managerial impact on workplace safety. [9][10][11][12] Within the literature on healthcare, there are non-empirical articles providing propositions and descriptions on managerial attitudes and efforts to improve safety and quality. This literature, made up of opinion articles, editorials and single participant experiences, present an array of insightful suggestions and recommendations for actions that hospital managers should take to improve the quality of patient care delivery in their organisation. [13][14][15][16][17] However, researchers have indicated that there is a limited evidence-base on this topic. [18][19][20][21] Others highlight the literature focus on the difficulties of the managers' role and the negative results of poor leadership on quality improvement (QI) rather than considering actions that managers presently undertake on quality and safety. [22,23]   How much time is spent by hospital managers on quality and safety and its improvement?
What are the managerial activities that relate to quality and safety and its improvement?
How are managers engaged in quality and safety and its improvement?
What impact do managers have on quality and safety and its improvement?
How do contextual factors influence the managers' role and impact on quality and safety and its improvement?  1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58   following dimensions: safe, effective, patient-centred, timely, efficient and equitable. [4] They define patient safety simply as "the prevention of harm to patients", [24] and AHRQ define it is "freedom from accidental or preventable injuries produced by medical care." [25] Literature was searched for all key terms associated with quality and patient safety to produce an allencompassing approach. A manager was defined as an employee that has subordinates, oversees staff, is responsible for staff recruitment and training, and holds budgetary accountabilities. Therefore, all levels of managers including Boards of managers were included in this review with the exception of clinical frontline employees, e.g. doctors or nurses, who may have taken on further managerial responsibilities alongside their work but do not have a primary official role as a manager. Those that have specifically taken on a role for quality of care, e.g. the modern matron, were also excluded. Distinction between senior, middle and frontline management were as follows: senior management hold Trust-wide responsibilities; [26] middle managers are in the middle of the organisational hierarchy chart and have one or more managers reporting to them; [27] frontline managers are defined as managers at the first level of the organisational hierarchy chart who have frontline employees reporting to them. Board managers include all members of the Board. Although, there are overlaps between senior managers and Boards (for example Chief Executive Officers (CEOs) may sit on hospital Boards), we aim to present senior and Board level managers separately due to the differences in their responsibilities and position. Only managers that would manage within or govern hospitals were included, with the exclusion of settings that solely served mental health or that comprised solely of non-acute care community services (in order to keep the sample more homogenous). The definition of 'Role' focused on actual engagement, time spent and activities that do or did occur rather than those recommended that should or could occur.

Screening
Three reviewers (AP, AR and DG) independently screened the titles and abstracts of the articles for studies that fit the inclusion criteria. One reviewer (AP) screened all 15,447 articles, while two additional reviewers screened 30% of the total sample retrieved from the search strategy: AR screened 20% and DG screened 10%. On testing inter-rater reliability, Cohen's kappa correlations showed low agreement between AR and AP (K=0.157, P<0.01) and between DG and AP (K=0.137, P<0.00). [29] However, there was a high percentage of agreement between raters (95% and 89% respectively), which reveals a good inter-rater  [30,31] Discrepancies were resolved by discussion and consensus. The main inclusion criteria were that: the setting was hospitals; the population sample reported on were managers; the context was quality and safety; the aim was to identify the managerial activities/time/engagement in quality and safety. The full inclusion/exclusion criteria and screening tool can be accessed in the online Appendices 2-3. Figure 1 presents the numbers of articles included and excluded at each stage of the review process.
Four hundred and twenty-three articles remained for full text screening. One reviewer (AP) screened all articles and a second reviewer (AR) reviewed 7% of these. A moderate agreement inter-rater reliability score was calculated (K=0.615,P<0.001) with 73% agreement. The primary reoccurring difference in agreement was regarding whether the article pertained to quality of care, owing to the broad nature of the definition. Each article was discussed individually until a consensus was reached on whether to include or exclude.
Hand searching and cross-referencing were carried out in case articles were missed by the search strategy or from restriction of databases. One additional article was identified from hand searching, [32] totalling 19 articles included in the systematic review.

Data extraction & methodological quality
The characteristics and summary findings of the 19 included studies are presented in Table   1. This Table is a simplified version of a standardised template that was used to ensure consistency in data extracted from each article. Each study was assessed using a quality appraisal tool developed by , [34] which comprised of two checklists (qualitative and quantitative). Random included articles (32%) were scored by AW for  rating criteria. The total percentage scores for each study are presented in Table 1. All studies were included regardless of their quality scores. Some cumulative evidence bias may results from two larger datasets split into more than one study each. [35][36][37][38] Through a narrative synthesis, we aimed to maintain the original meanings, interpretations and raw data offered by the articles. [39] Rating Criteria to verify whether question or objective is sufficiently described

Yes
Is easily identified in the introductory section (or first paragraph of methods section).
Specifies (where applicable, depending on study design) all of the following: purpose, subjects/target population, and the specific intervention(s) /association(s)/descriptive parameter(s) under investigation. A study purpose that only becomes apparent after studying other parts of the paper is not considered sufficiently described.
Partial Vaguely/incompletely reported (e.g. "describe the effect of" or "examine the role of" or "assess opinion on many issues" or "explore the general attitudes"...); or some information has to be gathered from parts of the paper other than the introduction/background/objective section.

No
Question or objective is not reported, or is incomprehensible.

N/A Should not be checked for this question
Box 1 Example of a rating criteria for Kmet's quality assessment [34]

RESULTS
This results section provides an overview description of the reviewed studies and their key findings. The findings are considered under four main headings: managerial time spent on quality and safety; managerial quality and safety activities; managerial impact on quality and safety; and contextual factors related to managers' quality and safety role. The section ends with a proposed model to summarise the review findings.  examined frontline staff (e.g. clinical directorate managers and unit nurse managers). The settings of the study were mostly Trust or hospital-wide; a few articles were set in specific settings or contexts: elderly care, [40] evidence-based medicine, [41] staff productivity, [42] clinical risk management, [43] and hospital acquired infection prevention. [44] Two studies involved specific interventions, [45,46] and 7 studies concentrated specifically on QI rather than quality and safety oversight or routine. [35,40,[45][46][47][48][49] There were a mixture of 6 qualitative design (interviews or focus groups); 8 quantitative survey designs; and 5 mixmethods designs. All but one study employed a cross-sectional design [46]. [46] The primary outcome measure used in most studies was perceptions of managerial quality and safety practices. All reported participant perceptions and a majority presented self-reports, that is, either a mixture of self and peer reports, or self-reports alone. [41,43,45,46] Several studies asked participants about their own and/or other managers' involvement in regards to their specific quality improvement intervention or quality/safety issue. [40,41,[44][45][46][47] With some variations, the most common research design was to interview or survey senior manager/Board members (particularly Board chairs, presidents and CEOs) perceptions on the Board/senior managers' functions, practices, priorities, agenda, time spent, engagement, challenges/issues, drivers and literacy (e.g. familiarity of key reports) on quality and safety. [35][36][37][38][48][49][50][51] Five of these studies included objective process/outcome measures, such as adjusted mortality rates. [35,37,38,49,50] No other studies included clinical outcome measures.
Similarly to the qualitative studies, 7 quantitative studies did not fully describe, justify or use appropriate analysis methods. However, compared with the qualitative studies, the quantitative studies suffered more from sampling issues. Three studies had particularly small samples (e.g. n=35) and one had an especially low response rate of 15%. Subject characteristics were insufficiently described in 5 studies; in one case the authors did not state the number of hospitals included in data analysis. Several studies had obtained ordinal data but only presented percentages, and only one study appropriately controlled for confounding variables. Across all articles, all but 3 studies reported clear objectives and asserted conclusions clearly supported by the data.

Managerial time spent on quality and safety
The studies on Board level managers highlight an inadequate prioritisation of quality and patient safety on the Board agenda and subsequent time spent at Board meetings. Not all hospitals consistently have quality on their Board agenda, for example CEOs and chairpersons across 30 organisations reported that approximately a third of all Board meetings had quality on their agenda, [35] and necessary quality items were not consistently  1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58  59 12 and sometimes never addressed. [36] In all studies examining time spent on quality and safety by the Board, less than half of the total time was spent on quality and safety, [32,37,38,[48][49][50][51] with a majority of Boards spending 25% or less on quality. [32,38,45,[49][50][51] Findings imply that this may be too low to have a positive influence on quality and safety, as higher quality performance was demonstrated by Boards that spent above 20%/25% on quality. [49,50] Board members recognised that the usual time spent is insufficient. [48] However, few reported financial goals as more important than quality and safety goals, [32] and health system Boards only spent slightly more time on financial issues than quality. [51] Similar issues are noted by studies on frontline managers; specifically that they placed less time and importance on QI, [42] identified as the least discussed topic by clinical managers. [52] Managerial quality and safety activities A broad range of quality-related activities were identified to be undertaken by managers.
These are presented by the following three groupings: strategy-centred; data-centred and culture-centred.

Strategy-centred
Board priority-setting and planning strategies aligned with quality and safety goals were identified as Board managerial actions carried out in several studies. High percentages (over 80% in two studies) of Boards had formally established strategic goals for quality with specific targets, and aimed to create a quality plan integral to their broader strategic agenda. [32,37] Contrary findings however suggest that the Board rarely set the agenda for the discussion on quality, [37] did not provide the ideas for their strategies, [32] and were largely uninvolved in strategic planning for QI. [48] In the latter case, the non-clinical Board managers felt that they held "passive" roles in quality decisions. This is important considering evidence that connects the activity of setting the hospital quality agenda with  1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58  59  better performance in process of care and mortality. [38] Additionally, Boards that established goals in four areas of quality and publicly disseminated strategic goals and reported quality information were linked to high hospital performance. [35,38,50] Culture-centred Activities aimed at enhancing patient safety/QI culture emerged from several studies across organisational tiers. [44,47,48,53] Board and senior management's activities included encouraging an organisational culture of QI on norms regarding interdepartmental/multidisciplinary collaboration and advocating QI efforts to clinicians and fellow senior managers, providing powerful messages of safety commitment and influencing the organisation's patient safety mission. [47,53] Managers at differing levels focused on cultivating a culture of clinical excellence and articulating the organisational culture to staff. [44] Factors to motivate/engage middle and senior management in QI included senior management commitment, provision of resources and managerial role accountability. [40,46] Findings revealed connections between senior management and Board priorities and values with hospital performance and on middle management quality-related activities. Ensuring capacity for high quality standards also appears within the remit of management, as physician credentialing was identified as a Board managers' responsibility in more than one study. [38,48] From this review it is unclear to what degree Board involvement in the credentialing process has a significant impact on quality. [38,41] Data-centred Information on quality and safety are continually supplied to the Board. [51] At all levels of management, activities around quality and safety data or information were recognised in 6 studies. [35,38,43,45,47,53] Activities included collecting and collating information, [43] reviewing quality information, [35,38,53] using measures such as incident reports and infection rates to forge changes, [53] using patient satisfaction surveys, [35] taking corrective  1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58  59  action based on adverse incidents or trends emphasised at Board meetings, [38] and providing feedback. [43,47] The studies do not specify the changes made based on the data-related activities by senior managers; one study identified that frontline managers predominantly used data from an incident reporting tool to change policy/practice and training/education and communication between care providers. [45] However, overseeing data generally was found to be beneficial, as hospitals that carried out performance monitoring activities had significantly higher scores in process of care and lower mortality rates than hospitals that did not. [38] Managerial impact on quality and safety outcomes We have considered the associations found between specific managerial involvement and its affect on quality and safety. Here, we summarise the impact and importance of their general role. Of the articles that looked at either outcomes of management involvement in quality or at its perceived importance, 6 articles suggested that their role was beneficial to quality and safety performance. [32,35,38,40,49,53] Senior management support and engagement was identified as one of the primary factors associated with good hospital-wide quality outcomes and QI programme success. [35,38,40,49] Conversely, 6 articles suggest that managers' involvement (from the Board, middle and frontline) has little, no or a negative influence on quality and safety. [35,38,41,42,44,49] Practices that showed no significant association with quality measures included Board's participation in physician credentialing. [35,38] Another noted that if other champion leaders are present, management leadership was not deemed necessary. [44] Two articles identified a negative or inhibitory effect on evidence-based practices and staff productivity from frontline and middle managers. [41,42] 1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58  59

Contextual factors related to managers' quality and safety role
Most of the articles focussed on issues that influenced the managers' role or their impact, as opposed to discussing the role of the managers. These provide an insight in to the types of conditions in which a manager can best undertake their role to affect quality and safety.
Unfortunately it appears that many of these conditions are not in place.
Two studies found that a Board quality committee is a positive variable in quality performance, but that fewer than 60% had them. [38,50] Similarly, compensation and performance evaluation linked to executive quality performance was identified in 4 articles [35,37,38,49] and associated with better quality performance indicators, [38,49] but quality measures were insufficiently included in CEOs' performance evaluation. [35,37] The use of the right measures to drive QI was raised in relation to Board managerial engagement in quality [35] and to impact on patient care improvement, [51] yet, almost half of this sample did not formally adopt system-wide measures and standards for quality. To aid them in these tasks, evidence indicates the common use of QI measure tools, such as a dashboard or scorecard, [37,49,50] with promising associations between dashboard use and quality outcomes. [38,50] Other factors linked to quality outcomes include management-staff relationship/high interactions between the Board and medical staff when setting quality strategy, [49] and managerial expertise. Although a connection between knowledge with quality outcomes was not found, [38] high performing hospitals have shown higher self-perceived ability to influence care, expertise at the Board and participation in training programs that have a quality component. [50] Disappointingly, there is a low level of CEO knowledge on quality and safety reports, [35] possibly little Boardroom awareness on salient nursing quality issues, [36] and little practice identified to improve quality literacy for the Board. [32,37] There is however promise for new managers through relevant training at induction and by recruitment of those with relevant expertise. [32]

The Quality Management IPO Model
The input process output (IPO) model is a conceptual framework that helps to structure the review findings in a useful way, please see Figure 2. [54,55] 1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58  59  Perceptions of managers on managemen t Board practices in quality and safety • Less than half (43%) of Boards reported that they addressed quality and patient safety issues in all meetings • One-third of Boards spend 25% of their time or more on quality and patient safety issues • More than 80% of Boards have formally established strategic goals for quality with specific targets, but a majority of Board chairs indicate that their Boards did not provide the ideas for strategic direction or initiatives • Board chairs reported a low participation in education on quality and safety: 43% reported that all the Board members participated, 19% stated that more than half participated and 23% said it was less than a quarter of the Board • Most Board chairs (87%) reported Board member induction training on responsibilities for quality and safety, although almost a third (30%) reported few or no opportunities for education on this, 42% reported some opportunities and 28% reported many) • Approximately half (57%) of the Board chairs acknowledged recruitment of individuals that have knowledge, skills and experience in quality and patient safety onto the Board. A Board skills matrix included quality and safety as one of the competency areas • Over half (55%) of board chairs rated their board's effectiveness in quality and safety oversight as very/extremely effective and 40% as somewhat effective. 16 Self-reported perceptions of managers on their engagement in a QI programme Five elements deemed essential to middle manager engagement: • (1) Senior management commitment and leadership ( e.g. senior management provides strategic direction for QI plan) • (2) Provision of resources and opportunities for QI education and information dissemination (e.g. basic QI skills provided to all staff) • (3) Senior and middle manager role accountability (e.g.: senior managers and middle managers agree QI roles and expectations) • (4) Middle manager involvement in QI planning (e.g. senior and middle managers plan together)  1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  Self-reported perceptions of managers on managers' quality and safety practices

•
The majority of managers (91%) who received collated incident information used it to feed back to their own staff. 60% always fed back to staff, 28% sometimes did, 2% never did.

•
Of the Trusts that had written guidance on types of clinical incident to report, 80% of managers had general guidance and fewer (20%) had written speciality specific guidance. • 76% of managers reported information collation of clinical incidents. Of these, 59% were involved in data collection themselves  Perceptions of managers on managers' role in quality and safety • Few CEOs were willing to take the lead for transformation to a "culture of quality" • Board members were largely uninvolved in strategic planning for QI • In terms of the Board's quality functions, respondents largely agreed that physician credentialing was their critical responsibility • Non-physicians reported that they felt relegated to "passive" roles in decisions on quality and seemed reluctant to assume leadership roles in the quality domain.

DISCUSSION
Our review examined the role of managers in maintaining and promoting safe, quality care.
The existing studies detail the time spent, activities and engagement of hospital managers and Boards, and suggest that these can positively influence quality and safety performance.
They further reveal that such involvement is often absent, as are certain conditions that may help them in their work.
Evidence from the review promotes hospitals to have a Board quality committee, with a specific item on quality at the Board meeting, a quality performance measurement report and a dashboard with national quality and safety benchmarks along with standardised quality and safety measures. Outside of the Boardroom, the implications are for senior managers to build a good infrastructure for staff-manager interactions on quality strategies and attach compensation and performance evaluation to quality and safety achievements. For QI programmes, managers should keep in mind its consistency with the hospital's mission and provide commitment, resources, education, and role accountability. Literature elsewhere supports much of these findings, such as the use of quality measurement tools [21,56] better quality-associated compensation, a separate quality committee, [16,57] and has also emphasised poor manager-clinician relationships as damaging to patients and QI. [58,59] Some of the variables that were shown to be associated with good quality performance, such as having a Board committee, compensation/performance and adoption of system-wide related to changes made to improve quality and safety in healthcare. [62][63][64] In addition to this evidence, a few studies have specifically investigated the impact that hospital managers have on quality and safety (rather than examination of their role). These studies have shown senior managerial leadership to be associated with a higher degree of QI implementation, [65] promotion of clinical involvement, [66,67] safety climate attitudes, [68] and increased Board leadership for quality. [57]  Research on this area is particularly required to examine middle and frontline managers, to take into consideration non-managers' perceptions, and to assess senior managers' time    • There is a dearth of empirical evidence on hospital managerial work and its influence on quality of care.
• There is some evidence that Boards'/managers' time spent, engagement and work can influence quality and safety clinical outcomes, processes and performance.
• Some variables associated with good quality performance were lacking in study hospitals.
• Many Board managers do not spend sufficient time on quality and safety.
• There is a greater focus on the contextual issues surrounding managers' roles than on examining managerial activities.
• Research is required to examine middle and frontline managers, to take into consideration non-managers' perceptions, and to assess senior managers' time and tasks outside of the Boardroom. More robust methodologies with objective outcome measures would strengthen the evidence.
• We present a model to summarise the evidence-based promotion of conditions and activities for managers to best affect quality performance. There are no competing interests.
What are the managerial activities that relate to quality and safety and its improvement?
How are managers engaged in quality and safety and its improvement?
What impact do managers have on quality and safety and its improvement?
How do contextual factors influence the managers' role and impact on quality and safety and its improvement?  1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58  following dimensions: safe, effective, patient-centred, timely, efficient and equitable. [4] They define patient safety simply as "the prevention of harm to patients", [24] and AHRQ define it is "freedom from accidental or preventable injuries produced by medical care." [25] Literature was searched for all key terms associated with quality and patient safety to produce an allencompassing approach. A manager was defined as an employee that has subordinates, oversees staff, is responsible for staff recruitment and training, and holds budgetary accountabilities. Therefore, all levels of managers including Boards of managers were included in this review with the exception of clinical frontline employees, e.g. doctors or nurses, who may have taken on further managerial responsibilities alongside their work but do not have a primary official role as a manager. Those that have specifically taken on a role for quality of care, e.g. the modern matron, were also excluded. Distinction between senior, middle and frontline management were as follows: senior management hold Trust-wide responsibilities; [26] middle managers are in the middle of the organisational hierarchy chart and have one or more managers reporting to them; [27] frontline managers are defined as managers at the first level of the organisational hierarchy chart who have frontline employees reporting to them. Board managers include all members of the Board. Although, there are overlaps between senior managers and Boards (for example Chief Executive Officers (CEOs) may sit on hospital Boards), we aim to present senior and Board level managers separately due to the differences in their responsibilities and position. Only managers that would manage within or govern hospitals were included, with the exclusion of settings that solely served mental health or that comprised solely of non-acute care community services (in order to keep the sample more homogenous). The definition of 'Role' focused on actual engagement, time spent and activities that do or did occur rather than those recommended that should or could occur.

Screening
Three reviewers (AP, AR and DG) independently screened the titles and abstracts of the articles for studies that fit the inclusion criteria. One reviewer (AP) screened all 15,447 articles, while two additional reviewers screened 30% of the total sample retrieved from the search strategy: AR screened 20% and DG screened 10%. On testing inter-rater reliability, Cohen's kappa correlations showed low agreement between AR and AP (K=0.157, P<0.01) and between DG and AP (K=0.137, P<0.00). [29] However, there was a high percentage of agreement between raters (95% and 89% respectively), which reveals a good inter-rater  [30,31] Discrepancies were resolved by discussion and consensus. The main inclusion criteria were that: the setting was hospitals; the population sample reported on were managers; the context was quality and safety; the aim was to identify the managerial activities/time/engagement in quality and safety. The full inclusion/exclusion criteria and screening tool can be accessed in the online Appendices 2-3. Figure 1 presents the numbers of articles included and excluded at each stage of the review process.
Four hundred and twenty-three articles remained for full text screening. One reviewer (AP) screened all articles and a second reviewer (AR) reviewed 7% of these. A moderate agreement inter-rater reliability score was calculated (K=0.615,P<0.001) with 73% agreement. The primary reoccurring difference in agreement was regarding whether the article pertained to quality of care, owing to the broad nature of the definition. Each article was discussed individually until a consensus was reached on whether to include or exclude.
Hand searching and cross-referencing were carried out in case articles were missed by the search strategy or from restriction of databases. One additional article was identified from hand searching, [32] totalling 19 articles included in the systematic review.

Data extraction & methodological quality
The characteristics and summary findings of the 19 included studies are presented in Table   1. This Table is a simplified version of a standardised template that was used to ensure consistency in data extracted from each article. Each study was assessed using a quality appraisal tool developed by , [34] which comprised of two checklists (qualitative and quantitative). Random included articles (32%) were scored by AW for  rating criteria. The total percentage scores for each study are presented in Table 1. All studies were included regardless of their quality scores. Some cumulative evidence bias may results from two larger datasets split into more than one study each. [35][36][37][38] Through a narrative synthesis, we aimed to maintain the original meanings, interpretations and raw data offered by the articles. [39] Rating Criteria to verify whether question or objective is sufficiently described

Yes
Is easily identified in the introductory section (or first paragraph of methods section).
Specifies (where applicable, depending on study design) all of the following: purpose, subjects/target population, and the specific intervention(s) /association(s)/descriptive parameter(s) under investigation. A study purpose that only becomes apparent after studying other parts of the paper is not considered sufficiently described.
Partial Vaguely/incompletely reported (e.g. "describe the effect of" or "examine the role of" or "assess opinion on many issues" or "explore the general attitudes"...); or some information has to be gathered from parts of the paper other than the introduction/background/objective section.

No
Question or objective is not reported, or is incomprehensible.

N/A Should not be checked for this question
Box 1 Example of a rating criteria for Kmet's quality assessment [34]

RESULTS
This results section provides an overview description of the reviewed studies and their key findings. The findings are considered under four main headings: managerial time spent on quality and safety; managerial quality and safety activities; managerial impact on quality and safety; and contextual factors related to managers' quality and safety role. The section ends with a proposed model to summarise the review findings.  examined frontline staff (e.g. clinical directorate managers and unit nurse managers). The settings of the study were mostly Trust or hospital-wide; a few articles were set in specific settings or contexts: elderly care, [40] evidence-based medicine, [41] staff productivity, [42] clinical risk management, [43] and hospital acquired infection prevention. [44] Two studies involved specific interventions, [45,46] and 7 studies concentrated specifically on QI rather than quality and safety oversight or routine. [35,40,[45][46][47][48][49] There were a mixture of 6 qualitative design (interviews or focus groups); 8 quantitative survey designs; and 5 mixmethods designs. All but one study employed a cross-sectional design [46]. [46] The primary outcome measure used in most studies was perceptions of managerial quality and safety practices. All reported participant perceptions and a majority presented self-reports, that is, either a mixture of self and peer reports, or self-reports alone. [41,43,45,46] Several studies asked participants about their own and/or other managers' involvement in regards to their specific quality improvement intervention or quality/safety issue. [40,41,[44][45][46][47] 1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58  59 1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58  59 12 and sometimes never addressed. [36] In all studies examining time spent on quality and safety by the Board, less than half of the total time was spent on quality and safety, [32,37,38,[48][49][50][51] with a majority of Boards spending 25% or less on quality. [32,38,45,[49][50][51] Findings imply that this may be too low to have a positive influence on quality and safety, as higher quality performance was demonstrated by Boards that spent above 20%/25% on quality. [49,50] Board members recognised that the usual time spent is insufficient. [48] However, few reported financial goals as more important than quality and safety goals, [32] and health system Boards only spent slightly more time on financial issues than quality. [51] Similar issues are noted by studies on frontline managers; specifically that they placed less time and importance on QI, [42] identified as the least discussed topic by clinical managers. [52] Managerial quality and safety activities A broad range of quality-related activities were identified to be undertaken by managers.
These are presented by the following three groupings: strategy-centred; data-centred and culture-centred.

Strategy-centred
Board priority-setting and planning strategies aligned with quality and safety goals were identified as Board managerial actions carried out in several studies. High percentages (over 80% in two studies) of Boards had formally established strategic goals for quality with specific targets, and aimed to create a quality plan integral to their broader strategic agenda. [32,37] Contrary findings however suggest that the Board rarely set the agenda for the discussion on quality, [37] did not provide the ideas for their strategies, [32] and were largely uninvolved in strategic planning for QI. [48] In the latter case, the non-clinical Board managers felt that they held "passive" roles in quality decisions. This is important considering evidence that connects the activity of setting the hospital quality agenda with  1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58  59  capacity for high quality standards also appears within the remit of management, as physician credentialing was identified as a Board managers' responsibility in more than one study. [38,48] From this review it is unclear to what degree Board involvement in the credentialing process has a significant impact on quality. [38,41] Data-centred Information on quality and safety are continually supplied to the Board. [51] At all levels of management, activities around quality and safety data or information were recognised in 6 studies. [35,38,43,45,47,53] Activities included collecting and collating information, [43] reviewing quality information, [35,38,53] using measures such as incident reports and infection rates to forge changes, [53] using patient satisfaction surveys, [35] taking corrective  1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58  59 14 action based on adverse incidents or trends emphasised at Board meetings, [38] and providing feedback. [43,47] The studies do not specify the changes made based on the data-related activities by senior managers; one study identified that frontline managers predominantly used data from an incident reporting tool to change policy/practice and training/education and communication between care providers. [45] However, overseeing data generally was found to be beneficial, as hospitals that carried out performance monitoring activities had significantly higher scores in process of care and lower mortality rates than hospitals that did not. [38] Managerial impact on quality and safety outcomes We have considered the associations found between specific managerial involvement and its affect on quality and safety. Here, we summarise the impact and importance of their general role. Of the articles that looked at either outcomes of management involvement in quality or at its perceived importance, 6 articles suggested that their role was beneficial to quality and safety performance. [32,35,38,40,49,53] Senior management support and engagement was identified as one of the primary factors associated with good hospital-wide quality outcomes and QI programme success. [35,38,40,49] Conversely, 6 articles suggest that managers' involvement (from the Board, middle and frontline) has little, no or a negative influence on quality and safety. [35,38,41,42,44,49] Practices that showed no significant association with quality measures included Board's participation in physician credentialing. [35,38] Another noted that if other champion leaders are present, management leadership was not deemed necessary. [44] Two articles identified a negative or inhibitory effect on evidence-based practices and staff productivity from frontline and middle managers. [41,42] 1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58  59

Contextual factors related to managers' quality and safety role
Most of the articles focussed on issues that influenced the managers' role or their impact, as opposed to discussing the role of the managers. These provide an insight in to the types of conditions in which a manager can best undertake their role to affect quality and safety.
Unfortunately it appears that many of these conditions are not in place.
Two studies found that a Board quality committee is a positive variable in quality performance, but that fewer than 60% had them. [38,50] Similarly, compensation and performance evaluation linked to executive quality performance was identified in 4 articles [35,37,38,49] and associated with better quality performance indicators, [38,49] but quality measures were insufficiently included in CEOs' performance evaluation. [35,37] The use of the right measures to drive QI was raised in relation to Board managerial engagement in quality [35] and to impact on patient care improvement, [51] yet, almost half of this sample did not formally adopt system-wide measures and standards for quality. To aid them in these tasks, evidence indicates the common use of QI measure tools, such as a dashboard or scorecard, [37,49,50] with promising associations between dashboard use and quality outcomes. [38,50] Other factors linked to quality outcomes include management-staff relationship/high interactions between the Board and medical staff when setting quality strategy, [49] and managerial expertise. Although a connection between knowledge with quality outcomes was not found, [38] 1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58  59

DISCUSSION
Our review examined the role of managers in maintaining and promoting safe, quality care.
The existing studies detail the time spent, activities and engagement of hospital managers and Boards, and suggest that these can positively influence quality and safety performance.
They further reveal that such involvement is often absent, as are certain conditions that may help them in their work.
Evidence from the review promotes hospitals to have a Board quality committee, with a specific item on quality at the Board meeting, a quality performance measurement report and a dashboard with national quality and safety benchmarks along with standardised quality and safety measures. Outside of the Boardroom, the implications are for senior managers to build a good infrastructure for staff-manager interactions on quality strategies and attach compensation and performance evaluation to quality and safety achievements. For QI programmes, managers should keep in mind its consistency with the hospital's mission and provide commitment, resources, education, and role accountability. Literature elsewhere supports much of these findings, such as the use of quality measurement tools [21,56] better quality-associated compensation, a separate quality committee, [16,57] and has also emphasised poor manager-clinician relationships as damaging to patients and QI. [58,59] Some of the variables that were shown to be associated with good quality performance, such as having a Board committee, compensation/performance and adoption of system-wide measures, were lacking within the study hospitals. There are also indications of the need to develop Board and senior managerial knowledge and training on quality and safety.  1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58  59  time is taken to consider quality of care matters at the highest level, an inference is that less attention will be paid to prevention and improvement of quality within the hospital. While the position that the item appears on the agenda is deemed of high importance, it is unimportant if the duration on this item is overly brief. In this vein, the inadequate time on quality spent by some may reflect their prioritisation on quality in relation to other matters discussed at the meetings or the value perceived to be gained from discussing it further. It might instead however be indicative of the difficulties in measuring time spent on quality by management.
Some of these studies provide us not necessarily with Board managers' time on quality and safety but their time spent on this at Board meetings. The two may not equate and time spent on quality may not necessarily be well-spent. [36] The emerging inference that managers greatly prioritise other work over quality and safety is not explicit, with further research required to identify what time is actually devoted and required from managers inside and outside of the Boardroom. Perhaps encouragingly, the more recent studies present more time spent on quality and safety than the earlier studies. Yet even the most recent empirical studies not included in our review conclude that much improvement is required. [60] This review presents a wide-range of managerial activities, such as public reporting of quality strategies and driving an improvement culture. It further highlights the activities that appear to affect quality performance. Priorities for Boards/managers are to engage in quality, establish goals and strategy to improve care, and get involved in setting the quality agenda, support and promote a safety and QI culture, cultivate leaders, manage resisters, plan ahead and procure organisational resources for quality. Again, much of the findings support the assertions made in the non-empirical literature. Above all, involvement through action, engagement and commitment have been suggested to positively affect quality and safety. [61] While researchers have stressed the limited empirical evidence showing conclusive connection between management commitment and quality, [21] some supporting evidence however can be unearthed in research that concentrates on organisational factors  26 related to changes made to improve quality and safety in healthcare. [62][63][64] In addition to this evidence, a few studies have specifically investigated the impact that hospital managers have on quality and safety (rather than examination of their role). These studies have shown senior managerial leadership to be associated with a higher degree of QI implementation, [65] promotion of clinical involvement, [66,67] safety climate attitudes, [68] and increased Board leadership for quality. [57] A clear case for the positive influence of management involvement with quality is emerging both from the findings of our review and Research on this area is particularly required to examine middle and frontline managers, to take into consideration non-managers' perceptions, and to assess senior managers' time  1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58  59  conduct studies on this topic. There is an over-reliance on perceptions across the studies, which ultimately reduces the validity of the conclusions drawn from their findings. As most of the study findings relied on self-reports, social desirability may have resulted in exaggerated processes and inflated outputs. Although, encouragingly, one of the included studies found that managers that perceived their Boards to be effective in quality oversight were from hospitals that had higher processes-of-care scores and lower risk adjusted mortality. The quality assessment scores should be viewed with caution; such scores are subjective and may not take into consideration factors beyond the quality assessment scale used. Due to the enormity of this review, the publication of this article is some time after the search run date. As there is little evidence published on this topic, we consider this not to greatly impact on the current relevance of the review, particularly as the literature reviewed spans almost  28 three decades. However, we acknowledge the need for an update of the data as a limitation of this review.

Conclusion
The modest literature that exists suggests that managers' time spent, engagement and work can influence quality and safety clinical outcomes, processes and performance. Managerial activities that affect quality performance are especially highlighted by this review, such as establishing goals and strategy to improve care, setting the quality agenda, engaging in quality, promoting a quality improvement culture, managing resisters, and procurement of organisational resources for quality. Positive actions to consider include the establishment of a Board quality committee with a specific item on quality at the Board meeting, a quality performance measurement report and a dashboard with national quality and safety benchmarks, performance evaluation attached to quality and safety, and an infrastructure for staff-manager interactions on quality strategies. However, many of these arrangements were not in place within the study samples. There are also indications of a need for managers to devote more time to quality and safety. More than one study suggest time spent by the Board should exceed 20-25%, yet the findings expose that certain Boards devote less time than this. Much of the content of the articles focused on such contextual factors rather than on the managerial role itself; more empirical research is required to elucidate managers' actual activities. Research is additionally required to examine middle and frontline managers, non-manager perceptions, and to assess senior managers' time and tasks outside of the Boardroom. We present the IPO model to summarise the evidence-based promotion of conditions and activities in order to guide managers on the approaches taken to influence quality performance. More robust empirical research with objective outcome measures could strengthen this guidance. • There is a dearth of empirical evidence on hospital managerial work and its influence on quality of care.
• There is some evidence that Boards'/managers' time spent, engagement and work can influence quality and safety clinical outcomes, processes and performance.
• Some variables associated with good quality performance were lacking in study hospitals.
• Many Board managers do not spend sufficient time on quality and safety.
• There is a greater focus on the contextual issues surrounding managers' roles than on examining managerial activities.
• Research is required to examine middle and frontline managers, to take into consideration non-managers' perceptions, and to assess senior managers' time and tasks outside of the Boardroom. More robust methodologies with objective outcome measures would strengthen the evidence.

3
exp Health Facility Administrators/ or Governing board/ or (Manager$1 or Matron$1 or CEO$1 or executive$1 or director$3 or board$1 or middle management or senior management or lower management or frontline management or leader$4 or president$1 or head of department$1.ti,ab. or department head$1.ti,ab. or head of nursing or administrator$1 or healthcare administration or (chief adj4 officer$1) or (chief adj4 nurs$) or (chief adj4 operation$1) or (chief adj4 service$1) or chief of staff).ti,ab.  1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58  59  Organization/ or Health$.ti,ab.)) or exp Health Care Quality/ or clinical effectiveness/ or incident report/ or nursing outcome/ or performance measurement system/ or quality of nursing care/ or length of stay/ or hospital readmission/ or evidence based medicine/ or exp outcome assessment/ or quality control/ or medical audit/ or patient satisfaction/ or (patient centred care or length of stay or customer satisfaction or patient satisfaction or medical audit or clinical audit or clinical effectiveness or performance measurement or outcome assessment or process assessment or guideline adherence or compliance to practice$1 or benchmarking or patient centred care or incident report$ or infection control or patient readmission or evidence based medicine or (evidence based adj2 practice) or waiting times or patient experience or complaints or target$1 or clinical excellence or service excellence or quality).ti,ab.

2
exp Hospital/ or Health care organization/ or (hospital$ or secondary care or acute care or health care organi*ation$1 or healthcare organi*ation$1 or infirmar$).ti,ab.

HMIC SEARCH
Step Search strategy formula 1 ((exp Safety/ or Safe$.ti,ab. or Err$.ti,ab. or Adverse.ti,ab.) and (Health Care Quality.mp. or Health$.ti,ab.)) or exp quality assurance in health services/ or patient centred care/ or hospital stay duration/ or exp consumer satisfaction/ or patient readmission/ or exp evidence based medicine/ or exp outcomes/ or benchmarking/ or (patient centred care or length of stay or customer satisfaction or patient satisfaction or medical audit or clinical audit or clinical effectiveness or performance measurement or outcome assessment or process assessment or guideline adherence or compliance to practice$1 or benchmarking or patient centred care or incident report$ or infection control or patient readmission or evidence based medicine or (evidence based adj2 practice) or quality improvement or waiting times or patient experience or complaints or target$1 or clinical excellence or service excellence or quality).ti,ab.

2
Acute Hospitals/ or hospital care/ or (hospital$ or secondary care or acute care or health care organi*ation$1 or healthcare organi*ation$1 or infirmar$).ti,ab.

Protocol and registration
5 Indicate if a review protocol exists, if and where it can be accessed (e.g., Web address), and, if available, provide registration information including registration number.
All protocol info in appendices Eligibility criteria 6 Specify study characteristics (e.g., PICOS, length of follow-up) and report characteristics (e.g., years considered, language, publication status) used as criteria for eligibility, giving rationale.

7
Information sources 7 Describe all information sources (e.g., databases with dates of coverage, contact with study authors to identify additional studies) in the search and date last searched.

6
Search 8 Present full electronic search strategy for at least one database, including any limits used, such that it could be repeated.
in appendices Study selection 9 State the process for selecting studies (i.e., screening, eligibility, included in systematic review, and, if applicable, included in the meta-analysis).

7
Data collection process 10 Describe method of data extraction from reports (e.g., piloted forms, independently, in duplicate) and any processes for obtaining and confirming data from investigators.

10
Data items 11 List and define all variables for which data were sought (e.g., PICOS, funding sources) and any assumptions and simplifications made.

6
Risk of bias in individual studies 12 Describe methods used for assessing risk of bias of individual studies (including specification of whether this was done at the study or outcome level), and how this information is to be used in any data synthesis.

DISCUSSION
Summary of evidence 24 Summarize the main findings including the strength of evidence for each main outcome; consider their relevance to key groups (e.g., healthcare providers, users, and policy makers).