Associations between Organisational Culture and Patient Outcomes: A Systematic Review

Table of Contents

Glossary…………………………………………………….

List of Abbreviations…………………………………………..

Abstract…………………………………………………….

1. Introduction………………………………………………

1.1. Background of Organisational Culture………………………….

1.1.1. Interpreting organisational culture……………………………

1.1.2. Cultural and organisational variables………………………….

1.1.3. Subcultures and Diversity………………………………….

1.1.4. Culture and performance………………………………….

1.2. Culture formation and transformation…………………………..

1.2.1. Managing culture……………………………………….

1.2.2. Managing competing cultures……………………………….

1.3. Measuring organisational culture………………………………

1.4. Rationale and Aims……………………………………….

2. Methodology………………………………………………

2.1. Overview of systematic review………………………………..

2.2. Minimising the risk of bias at the review level……………………..

2.3. Search Strategy………………………………………….

2.4. Scoping and search terms……………………………………

2.5. Searching Electronic bibliographical databases…………………….

2.6. Grey literature searches…………………………………….

2.7. Hand searched…………………………………………..

2.8. Data management………………………………………..

2.9. Research questions and inclusion criteria…………………………

2.9.1. Research questions………………………………………

2.9.2. Inclusion Criteria……………………………………….

2.9.2.1. Population…………………………………………..

2.9.2.2. Intervention………………………………………….

2.9.2.3. Comparators…………………………………………

2.9.2.4. Outcomes……………………………………………

2.10. Assessing the risk of bias…………………………………..

2.11. Methodological quality assessment using the Mixed Methods Appraisal Tool.

2.12. Data extraction………………………………………….

3. Results…………………………………………………..

3.1. Identification of records and study selection………………………

3.2. Study characteristics………………………………………

3.3. Patient outcome characteristics……………………………….

3.4. Organisational culture tool characteristics………………………..

3.5. Risk of bias within studies…………………………………..

3.6. Risk of bias across studies: language bias…………………………

4. Discussion………………………………………………..

5. Conclusion……………………………………………….

1.     Introduction

Priority on delivering high quality, safe healthcare has been a key policy objective for governments in much of the developed world for nearly two decades (Davies, Nutley & Mannion 2000; Dixon-Woods, McNicol & Martin, 2012). Many individuals and agencies connected with healthcare quality and performance, have accentuated the need for cultural change to be operated alongside structural, financial and procedural reforms (Mannion, Davies, & Marshall, 2005). Despite encouraging evidence of improving quality and safety, delivery of the aforementioned policies has been meek (Benning et al, 2011; Wachter, 2010). Patients across the developed world continue to suffer avoidable harm and substandard care (Wachter, 2010), England’s National Health Service (NHS) not being immune from these issues; the Bristol Heart Scandal, high rates of paediatric mortality after cardiac surgery at the Bristol Royal Infirmary, during 1984-1995 (Smith, 1998); and the Mid Staffs scandal, poor care and high mortality rates at the Mid Staffordshire NHS Foundation Trust (FT), between 2005-2009 (Healthcare Commission, 2009). The Kennedy inquiry report that examined the failings at Bristol Royal Infirmary concluded “the culture of healthcare which so critically affects all other aspects of the service which patients receive, must develop and change” (Kennedy, 2001, p. 277). Kennedy described the prevailing culture during the 1984-1995 as a ‘club culture’ that emphasised excessive power and influence around a core group of senior managers (Mannion et al, 2008). Similarly, the Francis inquiry report that examined failings at Mid Staffordshire NHS FT, identifying causes of organisational degradation as systemic, he saw the underlying errors as institutional and cultural in character (Dixon-Woods et al, 2013). Francis (2013) blamed an “insidious negative culture involving a tolerance of poor standards and a disengagement from managerial and leadership responsibilities” (p. 3).

Francis’s findings are disappointingly recognisable, examples of failures with other healthcare systems around the world: Canada, Netherlands and New Zealand have experienced similar crises in healthcare (Walshe & Shortell, 2004). Poor management systems, failure to respond to patient concerns and cultures of denial of uncomfortable information were some of the features that Walshe and Shortell identified. Walshe and Shortell (2004) recognised similarities with the characteristics of failures “The causes and characteristics of major failures in countries with different ways of organizing health care are remarkably similar” (p. 107). Nevertheless, managing organisational cultures within the NHS has been included as part of previous policy reforms dating back to 1984 (Mannion et al, 2008). The Griffiths’ report led to a number of resource management programmes and the development of general management in hospitals (Department of Health and Social Security [DHSS], 1983). These initiatives propelled internal market reforms that increased managerial control and accountability within the NHS fostering a competitive business culture throughout (Davies et al, 2000). A question to consider here is concerning the extent of failures and similarities seen at Mid Staffordshire and Bristol Royal Infirmary, whether their features are symptoms of a more widespread culture, particularly as other organisations within the NHS are open to the same institutional and regulatory nature. Especially as the role of organisational culture in improving quality management and performance is not limited to healthcare; in fact it is a focus within other industries (Lapina, Kairisa & Aramina, 2015). Lapina et al (2015) studied organisational culture of a university, they found that organisational culture is directly linked with effectiveness and performance – the stronger the organisational culture the more effective the organisation. Despite the importance of organisational culture and its role it has on quality and performance, there appears to be a paucity of research within healthcare.

This systematic review attempts to pull together what is known about organisational culture within healthcare and its associations with quality and patient outcomes. Systematic reviews aim to identify evaluate and summarise the findings of all relevant individual studies, whilst offering available evidence to become more accessible to decision-makers (Centre for Reviews and Dissemination [CRD], 2009). More so, where appropriate, combining the results of several studies provides a greater reliability and specific estimate of an intervention’s effectiveness than one study alone (CRD, 2009). This is authenticated due to the strict scientific design guidelines, with pre-specified, reproducible methods used, limiting the flaws seen in narrative reviews (Impellizzeri & Bizzini, 2012). Not only can systematic reviews layout what we know about a particular intervention, it also establishes where knowledge is lacking and can guide future research (Brown et al, 2006; Greens & Higgins, 2011; Petticrew, 2003). This introduction will continue with exploration of the origins and conceptions of organisational culture, the significance of this in healthcare, the formation and transformation of culture (OC), and the instruments and tools used for measuring organisational culture within a healthcare setting.

1.1. Background of Organisational Culture

Anthropological literature going back many decades has deeply embedded conceptions of ‘culture’ to indigenous people (Malinowski, 1922). Although the function of these conceptions to organisations originated in the United States, directly after the Second World War period, however came to favourable attention in the 1980s (Blau, 1955; Mannion et al, 2008). This period saw an emergence of best selling management books, which rooted the concept of organisational culture as paramount in the management of organisational performance (Deal & Kennedy, 1999; Peters & Waterman, 1982). Organisational culture has continued to be one of the key themes in organisational research as few contending ideas can compete its status; existing literature is plagued with special issues on content, impact and dynamics of culture, which has placed huge emphasis on managers to consider the implications for their organisations of its culture (Ogbonna & Harris, 2002).

Many of the ideas and themes on organisational culture during the 1980s and onwards however were not original; in fact, more moderately, it has been argued these concepts can be seen as a continuum of a trend that started early 20^th century (Mannion et al, 2008). Mannion argues the focus of study on organisational culture is recognised as a movement away from the ‘mechanistic perceptions’ of organisations, connected with concepts of scientific management. The perceptions of scientific management or ‘Taylorism’ as it became known, based on the contribution of Frederick Taylor; advocated that an in-depth understanding of the technical means of production, coupled with time study and financial incentives could lead to a significant improvement of an organisation’s efficiency (Dean, 1997). Some authorities view Taylor’s work as an exclusive interest solely based on the pursuit of an organisation’s efficiency with no regard for the human element involved (Sheldrake, 1996). This distortion however is contested by some who argue about the concept of Taylor’s work focused on a main concern being ‘cultural issues’, which encompassed pre-empted strands of human relations theory, as well as organisational literature (Parker & Bradley, 2000; Taska, 1992). Taska (1992) suggests that prevalent writings on organisational culture have been formed from Taylor’s theories and driving forces of scientific management.

Further influences throughout the 20^th century on organisational culture include the human relations theory, most noticeably the works of Elton Mayo in the 1930s (Mayo, 1924; see also Witzel & Warner, 2015). Mayo’s work highlighted the importance of using informal social constructions, when studying human behaviour within organisations (Parker & Bradley, 2000). More so, Wright (1994) suggests that Mayo used anthropological research methods and expertise, which opened the door to anthropology within organisational studies. The exploration of culture within organisations continued during the 1950s and 60s with cited works including, the study of banana time (Roy, 1959) and the changing culture of a factory (Jacques, 1951). Although between these times and up until 1980s studies in the area of organisational culture were meek (Parker & Bradley, 2000). There is broad acknowledgment that Pettigrew (1979) coined the term ‘organisational culture’, however Jordan (1994) argues, wider interest in the concept of organisational culture was due to three best selling books; Ouchi’s (1981) Theory Z, sometimes referred as ‘Japanese Management’ style; Peters and Watermans’ (1982) In Search of Excellence; and, Deal and Kennedy’s (2000) Corporate Cultures (Mannion et al, 2008). These books all provided the same message that in order for organisations to be successful, ‘cultures’ had to be the main focus (Jordan, 1994). Jordan (1994) highlights during these times American businesses were concerned by their Japanese competitors, as ‘organisational culture’ was seen as the success for Japanese organisations.

1.1.1.     Interpreting organisational culture

Despite the universal agreement that organisational culture exists and plays a fundamental role in shaping behaviour in organisations, there is little consensus on the definition and interpretation on organisational culture (Watkins, 2013). Van der Post, De Coning & Smit (1997) identified over 100 dimensions connected to organisational culture (see Table 1). Similarly, a critical review undertaken by Kroeber and Kluckhohn in 1963 identified 164 definitions of the term ‘culture’. This level of complexity exemplifies why there is no consensus, however without reasonable agreement it becomes difficult to understand the connections and other important elements that make up an organisation. This also prevents the ability to develop approaches to analyse and transform cultures (Watkins, 2013).

Table 1. 

Dimensions of Organisational Culture

Notwithstanding the differing views, Smircich (1983) distinguished two schools of thought. The difference between these schools is illustrated in Table 2. Firstly, there is a group of approaches that favour culture as something that an organisation is; this likens culture as a ‘root metaphor’ rather than being seen as something distinguishable from the organisation itself (Fieldman, 1986; Morgan, 1986; Wacker, 1995). Post-modern views on organisational culture sit within this school and are closest to the work of anthropologists, as they dispute the idea of organisations and their cultures as solid entities (Gregen, 1992). Lyn-Meek (1994) complements this idea and approaches culture as a ‘social emergent’ and not ‘created’ by the organisation but which emerges through social interaction and the transference of ideas, values and norms over time. Conversely, the second schools approach conceives culture as being something an organisation has; characteristics or variables of an organisation, which can be isolated, defined and influenced (Cummings & Schmidt, 1972; Schwartz & Davis, 1981; Schein, 2004). Smircich (1983) considers this concept to serve four main functions: providing stakeholders of an organisation with a sense of identity, it augments social system immovability, the concept enables the commitment to a better whole, and it functions as a sense-making device which can pilot and structure the behaviour of organisational stakeholders.

Table 2.

Culture as a Variable or a Root Metaphor

Importantly, these distinctions are vital for, if culture is something that an organisation has then this aspect can strategically be used to influence and direct the development of an organisation. Whereas, the idea that culture in an organisation is based on social entities offers less in terms of shaping change or supporting with management change, but provides understanding the processes of social construction. It seems clear that much of the advice targeted at organisations from management theorists assumes cultures are attributes as organisational variables that can be subject to influence (Davies et al, 2000). More so, the reforms we have seen in the NHS, since the Labour Government from 1997 have embodied this view (Davies et al, 2000). The differing views provides for much thought, however given the support from management theorists and current government reforms, this systematic reviews working assumption will be that culture is something that an organisation has, vis-à-vis, cultural attributes as organisational variables.

1.1.2.     Cultural and organisational variables

There is agreement amongst those who perceive culture as an organisational variable that shared beliefs, attitudes, values and norms of behaviour arises from what is shared between colleagues in an organisation (Davies et al, 2000). Langfield-Smith (1995) offers that organisational culture is reflected by a mutual agreement by people who see situations and events in commonality and distinctive ways. In lay terms organisational culture has been described as ‘the way things are done around here’, as well as ‘the way we think and act within organisations’ (Scahill, Harrison, Carswell & Babar, 2009; Schien, 2004). The most commonly cited definition of OC within healthcare literature is by Edgar Schein:

“A pattern of shared basic assumptions that the group learned as it solved its problems of external adaption and internal integration, that has worked well enough to be considered valid and, therefore, to be taught to new members as the correct way to perceive, think, and feel in relation to those problems.” (Schein, 2004, p. 17).

Deciphering the various aspects of OC can be overcome by clustering the various dimensions into different levels of culture. The most frequently used approaches to this are Schein’s typology (1989) distinction of artefacts, values, and basic assumptions. Artefacts the most visible and tangible, forms the top level of an organisation’s culture. These elements that may be readily manipulated include the physical, the environment, products, technology and patterns of behaviour within an organisation. The next level below artefacts is referred to as organisational values, and are seen operating at a conscious level. These values are seen as underlying and provide stimulus for behaviour, representing the standards and goals to which stakeholders within organisations attribute intrinsic worth (Child & Faulkner, 1998).

At the most basic level forms the basis for real cultural understanding, as it is the underlying assumptions that represent the unconscious; fundamental beliefs, values and perceptions impacting an individuals’ thinking and behaviour (Mannion et al, 2008). This level is different to the values level and is seen as unconscious, as these assumptions are so internalised to the stakeholder they drop out of one’s consciousness (Schein, 1989, 2004).

It is important and useful to see such differentiations bringing the aforementioned cultural levels within a healthcare context. Deep-rooted unconscious values and beliefs of a stakeholder may become difficult to change from external influence. The visible artefacts that represent manifestations of culture are more susceptible to manipulation (Mannion et al, 2008). Supporting this argument would be the Griffiths reforms in the 1980s, which put in place the principles of ‘general management’ into organisations in the NHS (Gorsky, 2013). These reforms were not successful in changing deeply embedded values and beliefs that underpinned clinical practice. Albeit, they did succeed in changing some manifestations of medical culture with the introduction of medical budgets and contracts (Gorsky, 2013). Furthermore, several of Griffiths’ concerns have exemplified in importance as aspects of health policy; including clinician involvement in improving performance and NHS management and finance (Gorsky, 2013).

There have been a number of adaptations to Schein’s work, which have tried to refine the typologies by subdividing the existing levels further. Ott (1989) splits artefacts into two, 1a art and technology; and 1b audible and visible patterns of behaviour. Hawkins (1997) goes further and divides this into five levels (see Table 3).

Table 3.

Hawkins Five Levels

Whilst Schein’s typology and the adaptations aforementioned offer a way of clustering various aspects of OC, they do not accommodate for the cultural diversity (sub- or micro-cultures), seen within organisations. Instead, there is an assumption that there is a predominant OC shared by all of the organisations stakeholders.

1.1.3.     Subcultures and Diversity

Martin (1992) argues that culture found within an organisation may not be as coherent as some may believe. Instead, Martin suggests rather than trying to identify commonality, examining the differences would be more rewarding. Martin (1992) proposes there are three perfectly valid perspectives to culture: Integration – culture is what people share, it is the glue that holds them together and consensus can be detected; Differentiation – culture is manifested by differences among subunits, OC is fraught with conflicts of interest and there is no consensus about what common culture exists; and Fragmentation – culture being ambiguous, culture is unknowable and not being an attribute of an organisation but the inherent nature of the organisation itself. Joanne Martin’s (1992) differentiation offers that some cultural attributes may exist across the whole organisation, whereas others may by more prevalent in sections of that organisation. This is an important notion, having different cultures evolving for instance, within different occupational or professional groups, with further subdivision of those groups with differing levels of power and control, which may alter over time. Such diversity can be likened to the domination of the medical culture within the National Health Service, with the fairly recent rise of the management culture (Jones & Dewing, 1997).  Still prevalent even twenty years on with the diversity of such professional groups; argued by Dixon-Woods et al (2013) that consistent achievement of high quality care is being challenged by unclear goals, and compliance-orientated bureaucratised management.

In 1994, Bolon and Bolon suggested that due to trends of increasing disintegration, rather than treating culture within an organisation as one singular entity, they recognised that a number of subcultures will exist. These trends still remain today with organisational subcultures being recognised as an important factor for motivation to transfer learning (Egan, 2008). These subcultures will occur within hospital settings since there is a subdivision with departments, specialties, and wards, which provides a harbinger for cultural diversity (Lok, Westwood & Crawford, 2005). A survey of 258 nurses conducted by Lok, Westwood & Crawford (2005), identified subcultures may be more of less prone to changing their values and beliefs, however a perceived organisational subculture has a strong relationship with commitment. It does seem evident that organisations can function with dissonant subcultures, whilst sharing some predominate cultural attributes. A prime example within the NHS would be how managers and doctor differ culturally, yet both groups inhabit a shared culture of medical independence. A paradoxical question raised by Davies et al (2000) is whether it is actually desirable that an organisation should pursue to achieve an integrated set of cultural attributes.  This supports the work of Bolon and Bolon, and is based on recognising that organisations will also gather cultural differences externally that inevitably will be at odds with the internal culture. Using this concept, the medical professional culture within a hospital will be influenced by national and even international trends. This dissention between prevailing professional cultures and the OC would appear to be a challenge. However, recognising that OC is itself a subculture within a larger set of ‘supra-cultures’, is important, especially when trying to explore culture, clarity to which level of culture being examined, is fundamental.

1.1.4.     Culture and performance

The introduction of culture within the domain of management was fostered by the notion that it may have an influence on organisational performance (Sackmann, 2011). Methods have been subsequently developed to aid understanding, assess and change corporate culture, with the aim to improve performance that would gain corporations a competitive advantage (Sackmann, 2011). Several scholars have cultivated this hope: Silverzweig and Allen (1976) observed changes in culture increase performance in six of eight firms they studied; Ouchi (1980) suggests unitary vision, with focus on humanistic values with a concern for employees, promoted financial success; and Peters and Waterman (1983) identified a correlation between superior performance and strong corporate culture. That said, a simple casual connection between cultural characteristics and success has not yet been demonstrated. More recently, Sackmann (2011) conducted a large review looking at 55 empirical studies on culture and performance. Overall her results suggested a contingency-type relationship between culture and performance, with an internal and external firm context.

Despite the inconclusive relationship between the two, OC still appears to be a predominate factor in understanding how organisations improve and perform. Many organisations public and private sector, place great emphasis in shaping their cultures in order to improve organisational ability (Sackmann, 2011). This is in keeping with the UK Government, who sees managing the culture is an avenue towards healthcare improvement:

“In the end, culture will trump rules, standards and control strategies every time, and achieving a vastly safer NHS will depend far more on major cultural change than on a new regulatory regime”. (Department of Health [DoH], 2014, p. 11).

1.2. Culture formation and transformation

Despite characters of culture enduring within an organisation, culture is more active and fluctuating, than being stationary (Davies et al, 2000). The activity may be due to rapid swings in the norms of an organisation, responding to crises, which may lead to changing the long-term direction. More so, new individuals joining an organisation may bring preceding expectations of culture; equally you would expect culture to be transferred to new individuals, unequivocally or implicitly by established staff. Old and new organisational features will also shape and articulate the culture; as well as influences from structures, routines and operational norms found within an organisation (Langfield-Smith, 1995). There may also be significant differences between open statements about cultural variables; this is pertinent within the NHS, especially as every trust has differing values and mission statements. Langfield-Smith (1995) adds that OC can have significant influence from factors outwith the organisation, particularly with the strong professional ethic and sense of purpose seen by health professionals. Davies (1999) agrees and offers that regulatory frameworks, public opinions and media all exercise influence, which heighten the implications for those trying to manage a cultural shift.

1.2.1.     Managing culture

There are a number of issues that need to be considered prior to attempting to transform culture within healthcare. Firstly, comprehensive and parallel change on the organisational variables previously covered, is unrealistic and not necessarily appropriate. An example within the NHS where a new quality strategy could develop on an existing cultural trait is that of justice and belief in the founding principles – no sick person should be denied medical aid because of a lack means. According to Davies et al (2000) centrality of patient care, using evidence based medicine and a willingness to examine quality issues are more contemporary values beginning to emerge. Davies continues that any strategy aiming to change culture should be selective, which allows for a balance between stability and renewal. This would allow identification of those cultural aspects that an organisation would want to keep and strengthen, and those that require adaptation.

Secondly, there is no clearly defined or specified cultural destination for the NHS or other healthcare providers within the UK. That being said in 1998, Shortell, Waters, Clarke and Budetti identified features of a new moral fabric for physicians. Their work recognised characteristics of the old expectations for physicians and brought them up to a new moral fabric. Also there is some evidence reading official policy (Five Year Forward View, 2014) explicates possible aspects of a desired cultural change. Albeit, seeing the true realisation of these shifts into day-to-day working remains to be seen (Davies et al, 2000). Davies argues that much work is still required on setting a clear vision for transforming the NHS in terms of cultural assumptions and values.

Thirdly, it is not easily possible to make cultural change from the top down by incitement. Although strategic approaches taking into account fears and motivations of staff at all levels would be more appropriate (Beer, Eisenstat & Spector, 1990).  More so, William, Dobson and Walters (1993) & Davies et al (1998) argue that any attempt to influence dimensions of culture need to be part of any wider improvement activities, including financial arrangements and strategic formulation. Within the NHS context this would include the likes of the ‘clinical governance project’ that is pivotal to cultural transformation; the National Institute for Clinical Excellence; and the National Performance Framework. According to Davies and Mannion (1999) the aforementioned would need to be integrated with clear and a consistent set of cultural values. Another necessity for an organisation is coherence, although may not be adequate to bring substantial change. Davies and Mannion (1999) are relating this to factors external to the organisation – professional bodies, external stakeholder groups and even the media, which may work against efforts of internal improvement.

1.2.2.     Managing competing cultures

According to Davies et al (2000) the key to achieving a desired transformation of culture within clinical governance falls to the critical relationship of the medical profession and service mangers. As these two professions have differing cultural roots the challenge would be to devise strategies of transformation that allows for a degree of cohesion between the professions. A valuable typology to classify the possible approaches to manage cultural diversity developed by Child and Faulkner (1998) includes four possible bases for accommodating competing cultures: Synergy – this base represents the fusing of both groups, taking the best elements of each to achieve the greatest possible fit between the two; Domination – this base acknowledges that fusing may prove difficult so instead accepts the right of dominance of a given groups’ culture; Segregation – this is based on seeking an acceptable balance between the either groups culture, maintaining separation rather than pursuing integration; and Breakdown – this occurs when one group seeks the aforementioned bases but fails to secure the agreement of the other group. So to aid cultural fit, the first three bases may hold the answer, however the fourth (breakdown) would likely lead to inactivity and organisational damage.

1.3. Measuring organisational culture

A range of tools and instruments designed to measure OC have been developed over the last few decades, which have been applied to industrial, educational and healthcare settings (Scott, Mannion, Davies & Marshall, 2003). These have been developed as some policymakers and academics have displayed further interest in the quantitative measurement of OC in order to determine its relationship with quality of care and performance (Davies et al, 2000; Scott, Mannion, Davies & Marshall, 2003b). Scott et al (2003) conducted a large study of the instruments and their usefulness for health service. This included instruments used in industry and education; in total, 13 instruments were reviewed, 9 of which have potential to be used in health care settings.  Scott and others identified the instruments would adopt either a typological approach; in which the results of the assessment would indicate one of more ‘types’ of OC; or a dimensional approach, which describes a culture by its view on a number of constant variables (Scott et al, 2003). Scott and others also identified that there is not an ‘ideal’ instrument to measure OC in healthcare, due to the limitations of each. Instead they found individuals might prefer choosing an instrument with either a positivist or constructivist approach, or even neither and use a qualitative approach such as observations or interviewing (Scott et al, 2003). To overcome such limitations, Ott (1989) used a multi-method approach, using two quantitative instruments together with an ethnographic approach. Albeit Ott had to delay his analysis of the quantitative data until he had spent some time observing the study to avoid contamination of data.   Furthermore an instrument that works well for one particular investigation may not be so effective for another study. This is due to the level of analysis, macro, meso or micro level; these levels will inevitably require different focus. Overall it is assumed that the purpose and context of cultural assessment would determine the choice of instrument required. Although, unlikely a single instrument will provide a valid, reliable and trustworthy assessment of an organisation’s culture, and therefore a multi-method approach would be more desirable.

1.4. Rationale and Aims

This background has shown evidence that cultures exist within organisations, which are unique to that organisation and carry subsequent sub cultures. Previous enquiries of healthcare failures have linked culture as a root cause of these failures. Anecdotally, the reviewer has experienced causes of poor patient quality to be also associated with organisational culture. There is belief that a healthy organisation is related to positive patient outcomes (Hesselin et al, 2013). There is an association between positive organisational cultures and high organisational performance. There are also tools available to measure culture within organisations, as well as managing competing cultures. However, as to date there is no systematic review that synthesises and appraises evidence-showing associations between healthcare organisations and patient outcomes. Therefore aim of this systematic review was to determine the associations between organisational culture and patient outcomes. This review will synthesise the evidence from previous researches, to determine if they possess the information regarding the below questions:

1) Are there associations between organisational culture and patient outcomes within a healthcare setting?

2) Do positive cultures within healthcare organisations lead to positive patient outcomes?

2.     Methodology

2.1. Overview of systematic review

To ensure validity, a systematic review is required to be based on the best available evidence (Ogilvie, Hamilton, Egan & Petticrew, 2005). According to the CRD (2009) systematic reviews should use a transparent reproducible methodology, a clearly stated set of objectives; using a search strategy that aims to identify all studies meeting the pre-specified eligibility criteria, which attempts to minimise bias; and assesses the methodological quality, synthesising and presenting the findings of individual studies in a systematic way.

This systematic review was produced and written in accordance with the Preferred Reporting Items or Systematic Reviews and Meta-Analysis (PRISMA) statement (Moher, Liberati, Tetzlaff & Altman, 2009). A completed PRISMA 21 index checklist of reporting items is provided in Appendix 1.

A single reviewer conducted all review processes undertaken as this systematic review was completed in fulfilment of an MSc dissertation.

2.2. Minimising the risk of bias at the review level

Reporting biases occur when the distribution of research findings is influenced by the nature and direction of results (Higgins & Green, 2011). There is evidence to suggest, statistically significant ‘positive’ results, have a higher likelihood of being published faster; a higher likelihood to be published in English; as well as being published multiple times in higher impact journals, which inevitably lead to higher citation. Importantly, any contribution that is made to the entirety of the evidence base within systematic reviews by studies that have produced non-significant results, are as important as studies with significant results (Higgins & Green, 2011).  It is essential to undertake a wide-ranging search strategy, which includes studies that report non-significant results to minimise the effects of reporting bias. To ensure this systematic review minimises reporting bias the search strategy has considered incorporating all relevant studies, irrespective of publication status. The caveat to this systematic review will be studies published in English, due to the review being undertaken by a single reviewer who only speaks English and has no resources available for translation. There are a number of types of reporting bias, which include: publication bias, time lag bias, multiple publication bias, location bias, citation bias, language bias, and outcome reporting bias. Please see Table 4 for types and definitions of reporting bias.

Table 4.

Types and Definitions of reporting bias

2.3. Search Strategy

According to Lefebvre, Manheimer and Glanville (2011) the principal aim of a search strategy is to achieve a high sensitivity, this will allow detection of all relevant articles; however caveated, Lefebvre and colleagues do assert this may yield in a low specificity, due to large quantities of unrelated articles may be identified. This needs to be taken into consideration, as there is only a single reviewer, and balance between sensitivity and specificity is required factoring in the resources available to them.

2.4. Scoping and search terms

An initial scoping literature search was conducted, which allowed the reviewer to get an understanding of the size of the published literature, the geographical spread of publications and the gaps in the literature. This initial review identified organisational culture and patient outcomes and their associations were not reported in the title or abstract of studies. Therefore, it was decided to use search terms aligned with the population and interventions of interest. The constructs: organisational culture, patient outcomes and healthcare were searched in isolation on three general databases (MEDLINE Complete, Cumulative Index to Nursing and Allied Health Literature [CINAHL] and Health Business Complete); all searches were conducted on the 3^rd June 2017. The idea of these searches was to measure the level of records available, which would provide an estimate of how manageable this review would be for a single reviewer. The Boolean operator ‘AND’ was added make the search more specific. Albeit, it quickly became evident that CINAHLs search returned an unmanageable number of records where as, MEDLINE and Health Business search retrieved very small numbers (Table 5).

Table 5.

Records retrieved from initial scoping review

From this finding and scoping the literature it became evident that the use of widespread keywords and phrases related to the ‘population’ and the ‘intervention’ would be required for an extensive search. This is advocated by The Cochrane Collaboration, with the use of the Boolean operator ‘OR’ (Lefebvre et al, 2011). Three lists of keywords and phrases related to the constructs of organisational culture, patient outcome and healthcare were created after initial scoping (Table 6).

Table 6.

Keywords and phrases related to the constructs: organisational culture, patient outcome and healthcare

* and $ symbolise truncation

The keywords and phrases related to the constructs in Table 5 were used in the search strategy. However, due to the high level results and specificity, search terms were reduced. Final search terms were: (work culture or organisation* culture) AND (Patient outcome* OR health outcome*) AND (Health organisation* OR hospital* OR health facility* OR acute care OR primary care OR health).

2.5. Searching Electronic bibliographical databases

To avoid bias in selecting studies it is recommended that a reviewer undertake a detailed search (McDonagh, Peterson, Raina, Chang & Shekelle, 2013). McDonagh and others go further and suggest hypothetical examples that can illustrate how selection of inclusion and exclusion criteria may introduce bias; however offer, dual review can mitigate bias (McDonagh et al, 2013). As one reviewer undertook this review, the aforementioned is difficult to achieve, however Lefebvre et al, (2011) recommend undertaking a comprehensive search using multiple databases. Consequently, the reviewer used multiple databases (Table 7) that were freely available via the University of Chester subscriptions.

To ensure a comprehensive return of searches the literature search was undertaken with a publication date of studies, since the inception of the databases to the start of June 2017, with one further update end of June 2017. To aid the reviewer limiters: full text, peer reviewed were used; source types: academic journal; gender: male and female; and no age barrier was used. Full details of searching electronic databases are provided in Appendices 1.

Table 7.

Descripts of electronic databases used

2.6. Grey literature searches

According the University of New England (n.d.) grey literature refers to research that is either unpublished or has been published in a non-commercial form. Examples of this include: Government reports, policy statements and issue papers, conference proceedings, theses and dissertations (University of New England, n.d.). Advocated by McDonagh et al, (2013) including grey literature in a systematic review is important, as it will help to reduce publication bias. Four sources used for grey literature search included: OpenGrey, an open electronic database system with access to 700,000 bibliographical references; OpenDOAR, an authoritative directory of academic open access repositories; BASE, a multi-disciplinary search engine to scholarly internet resources; and, Google Scholar, a freely accessible web search engine that indexes full text or metadata of scholarly literature.

2.7. Hand searched

The International Journal for Quality in Healthcare was hand searched, as this journal was not indexed in the electronic bibliographical databases. Hand searching records that are not retrieved by search terms can aid with reducing publication bias (CRD 2009, p. 18). Online access to this journal was available, which was accessible via the subscription held by the University of Chester. Only full-text articles were fully reviewed, however hand searching of article abstracts and titles were undertaken, for their appropriateness in this review.

2.8. Data management

All data collected from the electronic bibliographical electronic database, grey literature searches and hand-searched documents were accurately manually recorded in to a tabulated excel document.  On completion of inputting data a review was undertaken, any duplication at that point was removed.

2.9. Research questions and inclusion criteria

2.9.1.     Research questions

Are there associations between organisational culture and patient outcomes within a healthcare setting?

Do positive cultures within healthcare organisations lead to positive patient outcomes?

The aforementioned research questions will be addressed by a systematic review and synthesis of all literature relevant to associations between OC and patient outcomes, once the following inclusion criterion is met.

2.9.2.     Inclusion Criteria

The following criteria were used to aid identification of relevant studies to help with providing an answer to the research questions.

2.9.2.1.          Population

Globally, any adults or children of any age being patients that receive a form of nursing or medical healthcare.

2.9.2.2.          Intervention

A measured organisational cultural within a healthcare setting that has an effect on patient outcomes.

A measured positive organisational culture within a healthcare setting that has a positive effect on patient outcomes.

The healthcare setting can be an organisation, hospital, and facility; within acute or primary care.

2.9.2.3.          Comparators

Was organisational culture associated with patient outcomes positive or negative? AND, was the organisational culture seen as positive?

2.9.2.4.          Outcomes

Studies are required to include a measure of organisational culture and patient outcomes. Other inclusion includes: English language only.

To be included in the study all of the aforementioned criteria had to be met, if any of the studies failed the criteria then they were excluded.

2.10.                   Assessing the risk of bias

According to Hartling et al, (2009) and Higgins and Green (2011), there is confusion between the terms quality assessment and the risk of bias, as they are sometimes seen as a common alternative and can be used interchangeably. Higgins and Green (2011) define risk of bias as a systematic error or deviation from the truth, in results or interferences, which is interchangeable with internal validity. The Cochrane Collection (2005) refers to internal validity as the extent to which the design and conduct of a study are likely to have prevented bias. Whereas quality assessment is seen as the extent to which all aspects of a study’s design and conduct can be shown to protect systematic bias, non-systematic bias and inferential error (Lohr, 2004). To reduce confusion and simplify for the reader, this review sees risk of bias being the way a study has been planned and directed; and, quality assessment refers to all other characteristics of quality, which includes satisfactoriness of reporting.

To assist the reviewer there are tools available for assessing the risk of bias, some though, have questionable clarity and many variations in actually classifying quality, which are no longer recommended by the Cochrane Collaboration (Higgins & Green, 2011). However, what the Cochrane collaboration do recommended to use is The Cochrane Collaboration’s tool for assessing risk of bias (Higgins, Altman & Sterne, 2011). Developed to incorporate seven areas: generation of the allocation sequence; concealment of the allocation sequence; blinding; attrition and exclusions; other generic sources of bias; biases specific to the trial design; and biases that might be specific to a clinical specialty (Higgins et al, 2011). Not without its critics, Armijo-Olivo, Stiles, Hagen, Biondo and Cummings (2012) and Hartling et al, (2013) suggest some of the domains in the tool are too subjective, which leads to associations with low inter-rater reliability.  Hartling et al (2009) also offers a pre-requisite of in-depth training is required to ensure effective use of the tool, with users having previous experience of assessing risk of bias.  The Cochrane Collaboration tool for assessing risk of bias was not considered appropriate due to the reviewer’s lack of experience. More so the literature search results did not yield any randomised trials, and therefore the tool would be an inappropriate method to assess risk of bias.

A checklist that was initially considered was the Downs and Black Checklist (1998). This instrument assesses the methodology of both randomised and non-randomised health care interventions. The checklist had high internal consistency (KR-20: 0.89), although a lower index for external validity (KR-20: 054); with little difference between its performance of randomised and non-randomised trials (Downs & Black, 1998). The checklist contains 27 questions spilt across 5 sections: study quality; external validity; study bias; confounding and selection bias and; power of the study. However the reviewer identified that a number of studies within the literature search were observational and mixed methods; therefore Downs and Black checklist would not be appropriate as it assesses methodology in interventional studies.

O’Cathain, Murphy and Nicholl, (2008) argues that although there is a wide availability of critical appraisal tools for quantitative and qualitative research, there has not been a consensus on quality criteria for mixed methods research.  Appraising the quality of studies that have different methods is challenging and previously quality checklists may have required adaptation based on the type of studies included in a review (CRD, 2009). Albeit, mixed study reviews are becoming increasingly more valid and as such appraisal tools that provide quality appraisal for quantitative, qualitative and mixed method studies are recommended (Pluye et al, 2011). The Mixed Methods Appraisal Tool (MMAT) is a checklist that was developed to provide the quality appraisal for studies included in systematic mixed study reviews (National Collaborating Centre for Methods and Tools, 2015). MMAT appraises and describes the methodological quality for three domains: mixed; qualitative and quantitative (three sub-domains: randomised controlled, non-randomised, and descriptive) (Pluye et al, 2011). Two evaluations of the MMAT’s reliability have been completed, which showed an agreement between reviewers was fair to perfect on MMAT criteria (Pace et al, 2012; Souto et al, 2014). Further results from Pace et al (2012) showed an overall quality score of appraisal studies was substantial. As no previous systematic reviews have been undertaken looking at organisational culture and patient outcomes, there is no consensus of the most appropriate quality assessment tool to use. Therefore the reviewer decided that the MMAT was most appropriate for this systematic review (Appendix 2) given the mixed study reviews identified in the literature search.

2.11.                   Methodological quality assessment using the Mixed Methods Appraisal Tool

All studies the met the eligibility inclusion criteria were included in the final review and were subsequently rated for methodological quality using MMAT (Appendix 3). The Mixed Methods Appraisal Tool (2011) is comprised two parts, part 1: contains two screening questions, and 19 items corresponding to five types of studies; qualitative research (n=4), randomised controlled trials (RCTs) (n=4), non-randomised studies (NRS) (n=4), quantitative descriptive studies (QDS) (n=4), and mixed methods studies (n=11), of which; 4 items for the qualitative component; 4 for the quantitative component (RCT or NRS or QDS) and, 3 specific items for the mixed methods component. Part 2 contains a tutorial to assist the appraiser in answering the items related to the five types of studies.

2.12.                   Data extraction

Data extraction involves the transfer of information from the primary study reports to the synthesised data set (Kahn et al, 2012). Argued by Kahn and colleagues as a ‘subjective’ process, which is prone to error; they recommend a specially designed data extraction form. Noyes and Lewin (2011) offer several different approaches to extracting data from included studies, which include: inclusive or selective extraction of qualitative findings; extracting only a limited core set of items or extracting a wider set of items, or; using a theoretical framework to guide data extraction. Taking the aforesaid into consideration the data extraction form was created using the inclusion criteria developed for the literature review (population, intervention, comparators and outcomes). Therefore, the data extraction form was applied to all studies that met the full inclusion criteria. A copy of the data extraction form is attached in Appendix 4.

3.     Results

3.1. Identification of records and study selection

3.2. Study characteristics

3.3. Patient outcome characteristics

3.4. Organisational culture tool characteristics

3.5. Risk of bias within studies

3.6. Risk of bias across studies: language bias

4.     Discussion

5.     Conclusion