Discursive Activity in the Boardroom

Introduction

This article examines competing discourses within a set of organizational written texts (the minutes of board meetings). Specifically it explores how these organizational texts of national (Welsh), public sector, health care providers (Local Health Boards; LHBs) took account of the discourses of a global institution, Patient Safety, when faced with immediate and competing discourses from within the organizations. This took place at a time of national organizational reconfiguration with 21 organizations reconfigured into seven LHBs.

The understanding in this article is that, in keeping with Organizational Discourse Analysis (ODA), discourses have constituting effects (Alvesson & Karreman, 2011a, 2011b; Hall, 2001; Ricoeur 1971, 1981, 1991; Rosenau, 1992) including the establishment of institutions (Phillips, Lawrence, & Hardy, 2004) and deinstitutionalization (Maguire & Hardy, 2009); the privileging and deprivileging of subject positions (Oakes, Townley, & Cooper, 1998; Potter & Wetherell, 1987), and identities (Hardy, Lawrence, & Grant, 2005); and the legitimization and delegitimization of bodies of knowledge (Alvesson & Karreman, 2011a; Boje, Oswick, & Ford, 2004).

Organizational discourse analysis is complex and nuanced but the generally accepted principles are that it provides a framework for understanding organizational and interorganizational phenomena (Alvesson & Karreman, 2000; Grant, Keenoy, & Oswick, 1998; Hardy & Phillips, 1999; Morgan & Sturdy, 2000; Mumby & Clair, 1997; Putnam & Fairhurst, 2001) and not only emphasizes the importance of linguistic phenomena but also sees language as fundamental to the construction of social reality (Chia, 1996; Gergen, 1999). There is no one definitive methodological approach in ODA and there are a variety of approaches (e.g., see Alvesson & Karreman, 2000; Chouiliaraki & Fairclough, 2010; Iedema, 2011; Jaworski & Coupland, 1999; Mumby, 2011; Putnam & Fairhurst, 2001; Wetherell, 2001; Woodilla, 1998; Wood & Kroger, 2000).

Phillips et al. (2004) demonstrated the role of texts in the process of institutionalization. Using discourse analysis they developed a model in which institutions are constructed through the production of texts. They note that while institutions do, indeed, have a behavioral component this is essentially transitory. Meanings are created from collections of texts, and from compilations of discourses, as individual texts are produced, distributed, and consumed. This continued production, distribution, and consumption of texts constitutes institutions. The texts describe and communicate actions, and information about actions, to others (Taylor, Cooren, Giroux, & Robichaud, 1996). Through this process actions are influenced and institutionalization occurs. Institutions are self-policing conventions (e.g., Douglas, 1986) that develop over time from past practices and understandings and result in taken-for-granted facts that shape future practice (e.g., Barley & Tolbert, 1997; DiMaggio & Powell, 1991; Jepperson, 1991; Meyer & Rowan, 1977; Zucker, 1977). Deviations from institutional practices are costly as sanctions—increased costs—are imposed by the institutional members (Jepperson, 1991; Lawrence, Winn, & Jennings, 2001). Costs include economic costs in the form of increased risk, cognitive costs as more thought is required (rather than automated practices), and social costs as legitimacy is reduced with reduced access to resources (Phillips, Lawrence & Hardy, 2000). Institutionalization is the social process through which individuals come to share a social reality (Scott, 1987, 2001). There is the caution that not all discourses result in the construction of institutions and not everything that is socially constructed is an institution. The distinguishing factor is that institutions are self-regulating and exercise control. However, there are challenges to the reification of institutionalization in relation to managerial control in postbureaucratic society (e.g., see Raelin, 2011; Spreitzer & Mishra, 1999; Styhre, 2008).

The board functions at one of the interfaces between the environment and the organization (e.g., He & Baruch, 2010), and must take account of competing external and internal discourses. The board is one of the “institutional manifestations” of the organization (Parsons, 1960). That is to say, the board is consciously engaged in institutionalizing processes as articulated by Parker (2007). These processes include legitimization and the adoption and inculcation of beliefs, and the approving of the routinization of practices. In relation to this, the board meetings, and related meetings, are ritualized and ceremonial organizational activities (e.g. see Islam & Zyphur, 2009; Parker, 2007).

The purpose of this project is to describe and critique the organizations’ worlds, their institutions, as they are revealed, re-presented, and re-created, in written text, within the field of health care. This is possible because the texts, in this instance, reveal these processes and worlds as well as constructing them.

The use of the term “revealed” does not, here, suggest an objective reality regardless of context. Rather, it is used to indicate that the traces of events recorded in the texts both reveal meaningful actions and are meaningful actions.

The question that developed for this article, in keeping with the purpose of the wider project, was as follows. What are the socially constructed organizational (social) worlds, or realities, that are revealed, and re-presented in the text in relation to a particular discourse? And, what are the wider theoretical and practical implications in terms of discourse, and the constructive properties of discourse? These questions are set in the context of public sector health care organizations in Wales responding to a specific national, systemwide, intervention (1000 Lives) which is in turn embedded in an industry level institution (Patient Safety) at a time of organizational reconfiguration.

The contribution made by this article is that it shows how an often neglected genre of organizational texts has important implications for organizational processes including institutionalization and deinstitutionalization at one of the key interfaces between the organization and the wider environment.

The article begins with a discussion of ODA and texts. Some consideration is then given to the extended “role” of the minutes and their constituting effects, especially the issue of recursiveness and cultural persistence (Zucker, 1977). The methodology section describes the context, the method used to extract data points from the text, and the consideration of the part-whole relationship. Following this the findings are presented and then a brief discussion.

An Approach to Organizational Discourse Analysis (ODA)

Some consideration was given to the principles both Foucaultian Discourse Analysis (FDA; Foucault, 1969/72; Kendall & Wickham, 1999; Willig, 1990), and Critical Discourse Analysis (CDA; Fairclough, 2009; van Dijk, 1993). Although these two approaches are often considered contradictory rather than complementary, they are both prevalent methodological and theoretical approaches in ODA (for a discussion on the conflict and complementarity of FDA and CDA see Graham, 2005; see also Chouliaraki & Fairclough, 2010).

What is common to both FDA and CDA, apart from considerations of power, is the importance of the relationship between language and other social processes; placing texts in context; and the statement, or even the word, having a constitutive effect. Thus units in the text have important effects and collections of similar and related units in the texts, which form networks of relations, have further effects. Where the two approaches are usually considered to differ is the FDA denies methodological hegemony, preferring instead a clear description of the way in which the research/analysis was undertaken, whereas CDA advocates a systematic approach and links to structuralism. Recent developments in hermeneutics readily allows for these two approaches, FDA and CDA, to be complimentary rather than contradictory or at least for them to be reconciled within a framework of multiple ontologies and epistemologies (see Cunliffe, 2011).

Reflections on the Principle of Constituting Effects of Discourse

Regardless of definition, it is accepted that discourse produces sets of culturally and historically situated concepts, which are “ideas, categories, relationships and theories through which we understand the world and relate to one another” (Hardy & Phillips, 1999, p. 3), and has an effect (Carabine, 2001; Parker, 1992). As a discourse becomes established the effect is to reify a social reality that is taken for granted (Phillips et al., 2004), and the assumption is that the more structured the discourse the greater the likelihood that it will lead to established practices (Phillips et al., 2000).

Discourse provides the language, knowledge, and a way of talking about a topic that is normalized and acceptable, and defines the cast of characters (the actors) who have normalizing roles within the discourse (Heracleous & Barrett, 2001). Discourses are constructed from, or emerge from, normative ideas and common sense and produce, and reproduce, ideas with connotations of morality, worth, and appropriate and acceptable behaviors (Barge & Oliver, 2003; Carabine, 2001; Knights, 1992; Townley, 1993). Characters achieve subject positions; those bureaucratic positions and legitimate categories of identity that warrant having a voice (Potter & Wetherell, 1987). Within any discourse there can only be a limited number of subject positions that have a legitimate voice with meaning and power (Hardy et al., 2005), and the right to speak and act, including the right to produce and distribute texts (Maguire, Hardy, & Lawrence, 2004). These subject positions are not fixed but change over time (Oakes et al., 1998).

Put another way, appropriately structured discourses with a high level of coherence and accessible meanings based on, or strongly related to, existing symbolic representations and with a strong relationship between the symbolic representations in the discourse, tend to effect the behavior of actors (either as intended, or unintended, or both). That is to say, discourse enhances some possibilities more than others (Reed, 1998), and actors can bring about preferred outcomes by drawing on preferred discourses (Deetz & Mumby 1990). Similarly, actors will discount or discredit those discourses that have the potential to bring about unwanted outcomes (Grant et al., 1998; Maguire & Hardy, 2009).

Reflections on Text or the Inscription of Discourse

A text is something that can be read and it is accepted within the field of discourse analysis, and hermeneutics, that texts are not limited to written documents. Indeed, written documents are a special case of texts due to the use of language and the changes that take place between dialogical interchange (speech) and the inscription of language. Ricoeur (1971) explains the significance of this change as follows. The inscription of discourse is the writing down of meaning. With written discourse the author’s intention and the meaning cease to coincide. Thus the concern of authorial intentions in the inscription of minutes failing to represent the actual dialogical interchange is moot. Ricoeur explains thus: “But the text’s career escapes the finite horizon lived by its author. What the text says now matters more than what the author meant to say.” (p. 529). This is not taken to mean that the author’s intention is irrelevant, rather that the author can no longer control either the meaning or the construed intention. He continues that in writing the discourse is freed from both the author and the dialogical situation and comes to project a world. That is, “a text breaks the ties of discourse to all the ostensive references” (p. 543), and this allows the text, the discourse in the text, to develop nonostensive references as an ontological dimension—new “worlds” (congruent with the construct of institutions) are created. However, the thrust of Ricoeur’s (1971, 1991) argument is that meaningful human action can also be read, which it can, and is therefore a text, which it is, but this text—meaningful human action—has similar properties to that of a written text in that the consequences are not under the control of the actor but escape the actor’s finite horizons, which is the property of any text. The point Ricoeur makes is that social science research relies on being able to read a text, following the wider definition of text (which is similar to Bruner’s (1991) requirement for narrative competence).

This has a particular relevance for the minutes of meetings as the legal status of a set of minutes is, generally speaking, one of a true and accurate record of the meeting in terms of structure, albeit it limited content, decisions, and actions. That is to say, as a text the minutes are a purported record of intentional and meaningful activity. Within an organization they are a special case of a text of body of texts (the texts presented to the board and the dialogical interactions of the board over successive meetings). Not only that, the minutes are by their very nature recursive. That is, the minutes of the board meeting at time t are re-presented, and reread, in the meeting at time t+1 and consensually agreed, sometimes by default, as a true and accurate reflection of the proceedings of the previous meeting at time t. Thus the minutes continually reinforce and propagate ideas, meanings, ways of thinking, what is taken for granted, beliefs, and so on; they are an institution (Czarniawska, 2004) and they have institutionalizing effects (Zucker, 1977). (For the importance of texts in maintenance, see also Law, 1994.) But the minutes, as any text does, escape their authors’ finite horizons and create new worlds. This is not, though, a static process occurring intermittently as the minutes are collated, even if by default, to form structured collections that can, and may, produce social categories and norms that shape the understandings and behaviors of actors (again, the minutes create a body of texts that form an institution and that have institutionalizing effects). The minutes also function to allow and disallow, and privilege and discredit, subject positions, bodies of knowledge, and further discourses as they escape the authors’ finite horizons—these are the constituting effects of language, discourse, and texts (again, the argument repeats).

Finally, with regard to the minutes, it seems reasonable to assume that board members will seek to have their discourses, their worlds of meaning, recorded in the minutes as they are all authors of the text and they presumably seek to maintain their institutional subject position. And, as Reed (2004), suggests, over time discourses are contested and preserved with the intentions of, for example, both the creation of new institutions and the preservation of the institutional status quo.

The Contexts

This section outlines the selected contexts. First, consideration is given to the history of Patient Safety and the linked 1000 Lives program in Wales. Second the 2009 reconfiguration of NHS Wales is referred to. This provides the context for the identification of the discourses of interest (or at least those discourses that were initially of interest).

Method

The analysis had six stages combining aspects of deconstruction, content analysis, and narrative analysis. (In this article deconstruction refers to an attempt to understand the text in relation to its context.) First, an archive was constructed from which a corpus was extracted (Fairclough, 1992). Second, the corpus was read, and reread, in depth with extensive note taking. Third, key segments—snippets—of text were extracted. Fourth, the selection of segments was revised using a computer search for key words within the corpus. Fifth, the snippets of text were sorted, resorted, and grouped to reconstruct discourses, narratives, and related events together with extensive memoing. Finally, a set of questions was used to interrogate the resulting reconstructions. These stages were interwoven and iterative rather than being completely linear as the author in turn iterated between inductive and deductive analysis. This iteration included returning to the corpus, and sometimes to the archive, to confirm observations made in the corpus.

The archive was built from the minutes of the LHBs’ board meetings, and the other documents presented at the meetings, collated from the LHBs web sites and organized by LHB and date. All documents were converted to text format as well as being kept in their original format. The LHBs web sites were also searched using the key terms “1000 Lives” and “Patient Safety” (with and without capitalization), and any additional documentation was collated. (These key terms were determined by the simple logic that the discourse/s of interest were 1000 Lives and Patient Safety and these terms were used for the search of the corpus.)

The corpus was selected from the archive. It consisted of agendas, minutes (of the board meetings and associated committees), and minutes from public meetings where the LHB was represented (e.g., AGMs and a small number of special meetings).

There were around nine sets of minutes per LHB, on average around 30 pages in length. The number of board meetings in the time period ranged from 7 to 12 (October 2009-December 2010). The mean number of documents reviewed was 28 per board over the time period. The range of analyzed documents per board was from 9 to 65.

The corpus was read in depth and passages of text relating to 1000 Lives and Patient Safety were extracted. Passages of text ranged from a single sentence to entire paragraphs or more. From these passages of text a narrative account of “events” within the text was reconstructed (Maguire, 2004). This account highlighted salient fragments or snippets of text that referred to 1000 Lives and/or Patient Safety. This provided a focus similar to what Fairclough describes as “moments of crisis” (Fairclough, 1992). Focusing on these moments and reading the text around them in depth highlighted conflicts and further, related, discourses that either complemented or contradicted the 1000 Lives and Patient Safety discourses, and discourses that subverted the 1000 Lives and Patient Safety discourses (this was made possible through an extensive understanding of a chosen wider context, in this case Patient Safety).

For example, 1000 Lives developed the intervention of Patient Stories, in this setting, to be presented as a 0 (zero) item on the agenda prior to the start of the board meeting. In a number of cases savvy people used this space as an opportunity to pitch for resources, and were awarded the resources. Most of the boards seemed to recognize this and took steps to prevent it, often by removing the space and modifying its use. Instances of this activity were noted and linked to 1000 Lives. This also served as an example for how other events might be linked in and noted.

It was this process that allowed competing discourses to be identified as in, for example, the concern raised by board members and members of the public that finance was being privileged over safety. This led to the identification of the board level Wales-wide discourses of Finance and Quality and Safety which were linked to 1000 Lives.

Passages of text ranged from a single sentence to entire paragraphs or more. Notes and memos were made relating to possible meanings and events and the activity recorded in the text. Possible implications and linked activities and events were noted. (This is similar to an events database, Van de Ven & Poole, 1990.)

Two groups of interest were constructed from this reading, based on their use of discourses and texts to construct social realities at the level of the board and to have these social realities inscribed in the minutes. The two groups were “Nursing” and “Chief Executives” with attendant discourses relating to, or in contradiction to, 1000 Lives and Patient Safety. These groups emerged naturally rather than being predetermined and were somewhat of a surprise. However, the seeming clarity of their position and their discourses prompted the author’s interest.

Following the reading and collation of sentences/fragments a simple automated word search using a text search program, Windows Grep (version 2.3.0.2403), was carried out for the terms “1000 Lives” and “patient safety” (capitalization not necessary). The 10 lines of text around each search term was captured and read and these passages of text were compared to the passages of text found in the manual search (see Figure 1). A total of 299 fragments were extracted from the minutes. Each fragment could be traced back its place, its context, in the original text. This process was repeated using the names of the Chief Executives in relation to the findings relating to the Chief Executives’ challenge given below.

OPEN IN VIEWER

Figure 1.

Three examples of fragments of text extracted using Windows Grep search for 1000 Lives from single board in the month of July 2010

The sets of sentences/fragments were sorted, ordered, compared, and contrasted to construct emerging themes, narratives, and re-presentations of discourses over time. They were grouped by month and common, or similar and related, fragments of text collated across the all the boards. These groups of fragments were already chronologically ordered and the emerging pattern over time was noted taking account of what was “talked about” (i.e., recorded in the text); and how the recorded discourse/s developed over time.

From these passages of text a narrative account of “events” within the text was reconstructed (Brown, 2004; Maguire, 2004). This account highlighted salient fragments or snippets of text that referred to 1000 Lives and/or Patient Safety. It was possible, from this in depth reading and the selection of fragments of text (which we can liken to fragments of stories, Boje, 1991), to identify and collate further discourses that either complemented or contradicted the 1000 Lives and Patient Safety discourses, and discourses that subverted the 1000 Lives and Patient Safety discourses (this was made possible through an extensive understanding of a chosen wider context, in this case Patient Safety; see Phillips & Hardy, 2002).

The identification of discourses was further facilitated throughout by posing a set of questions (see Prasad, 2002) derived from (a) the theoretical frameworks employed; and (b) wider questions of interest relating to the texts. (See Table 1.)

OPEN IN VIEWER

Table 1.

Questions Drawn From Theoretical Frameworks

Questions
Is the discourse external to the organization or internal to the organization; what are the ideas, truths, and so on; what effect is it having (how is it being used)
Are the discourses competing/combining?
What are the subject positions?
What is the legitimized “body of knowledge” underpinning the discourse, and what are the components of this body of knowledge?
Are texts being combined and what are the implications/effects?
How is the text (the minutes) being used; what is being recorded; what are the implications?

These questions provided a means for the researcher to ask questions of the corpus and of the text now compiled from the collated segments.

Summing up, the parts of the texts relating to the wider social discourses were extracted and deconstructed (Brown, 2004). This resulted in related sets of ordered snippets of text relating to the discourses. This was all carried out with an understanding of contexts, and especially the historical context of Patient Safety and 1000 Lives.

This process allowed for the reconstruction of a text-within-the-text (the corpus) as a “tissue of meaning which is symbolically significant for the reader (Parker, 1992)” (cited in, Parker, 1999, p. 3). This is in keeping with the identification of traces left by discourses in the texts (Parker, 1994), which in turn become further discourses as the reader reconstructs “meaningful totalities out of scattered events” (Ricoeur, 1981, p. 279; see also de-fragmentation: Alvesson & Karreman, 2011b, p. 41). Put another way, the experience of the researcher becomes a conscious effort to mimic, to some extent, the presumably unconscious experience of the readers (Prasad, 2002), albeit from a removed position, which is in keeping with Ricoeur’s discussion of the text as action and action as text (1971, 1981), and reading as a recovery of meaning (1981), and where we take a broad view of meaning.

The author was also conscious of their own institutional pressures to consider particular arguments, points of view, styles, perspectives, and so on. However, this article is an attempt to share a set of understandings, based on observations, that have further implications, both pragmatic and theoretical. It is not an attempt to argue for one particular point of view. For their own part, the author is a relatively experienced board member, and committee member, across a number of sectors, has been a company secretary, and has broad experience of both taking minutes and presenting minutes.

Findings

Although the analysis focused on text segments related to 1000 Lives and Patient Safety further discourses were clearly identifiable as the discourses occurred in proximity to each other in the texts in the corpus. It seemed that the discourse of 1000 Lives was challenged and its prominence at board level reduced over time.

The early symbolic meaning of 1000 Lives was related to that of Patient Safety and was embedded in a body of wider societal discourses relating to quality improvement, risk reduction, and systems improvement (IOM Report; Kohn et al. 1999/2000). What was lacking over time is concrete evidence that the program achieved the stated outcome of improving standards of care and reducing levels of harm incurred by patients receiving health care. The embedding of the discourses and the lack of evidence is mirrored in the minutes.

The minutes contain statements such as the board is “. . . committed to participating in the all Wales 1000 Lives patient safety campaign . . .” and there are press releases, from 1000 Lives, for at least two boards saying they are focusing on Patient Safety. A consistent statement across the boards LHBs is to fete 1000 Lives as a program that is being implemented for the “. . . first time in the world . . . ” and the principles and achievements should be “embedded.”

Support for PSIs emerging from 1000 Lives is strongly in evidence in the January 2010 minutes of Hillside LHB. Organizational members of this LHB claim to have developed the “SKIN” bundle, a bedsore (ulcer) prevention and treatment intervention, which has been “adopted” by 1000 Lives and “rolled out” (these claims are partially contested by Valley LHB, April 2010, who have their own intervention). One of the reported outcomes is a “culture change” regarding ulcers—previously it was taken for granted, by nurses, that pressure ulcers were inevitable and unavoidable. The 1000 Lives program is also credited with reducing levels of infection and improving discharges of patients from hospital.

Yet even at this early stage tensions emerge. The minutes record that a board member, at Western LHB, requested details of savings and achievements from 1000 Lives but did not receive them. The same board refers to the high cost of clinical negligence claims across Wales, especially midwifery in their geographical region, and the lack of incident reporting. The endoscopy unit at Valley LHB (February 2010) has strong failings, including being unsafe for recovery and is reported as being in the same position 10 months later. The minutes of Hillside LHB (December 2009) record that hand hygiene targets are linked to 1000 Lives but there has been no reduction in MRSA. In May 2010 the minutes record that the use of Hospital Standardized Mortality Rates (HSMR; see below) set by government have a high variation in coding across Wales ranging from 6% to 20%. Within Wales this data is not available for interorganizational comparisons and the 1000 Lives web site states, “The central Campaign office at Public Health Wales will only receive an all-Wales HSMR figure. It does not carry any individual Trust data or comparative data with England or any other countries” (1000 Livesb; HSMR is the Hospital Standardized Mortality Rate, a weighted measure of hospital deaths that seeks to account for risk of mortality from the point of admission to hospital). This comparative HSMR data is in the public domain for the remainder of the United Kingdom via the Dr. Fosters web site.

In addition to this, tensions and contradictions between discourses and occupational groups, in relation to quality and safety, quickly become apparent. The two prominent groups, with related discourses, are Nurses and Chief Executives.

The Chief Executives’ (CEs) challenge

Rather than challenging the discourse of 1000 Lives, the Chief Executives’ challenge was to institutionalize a “new” discourse of Finance. However, they faced a further challenge in as much as this discourse had to be somehow merged with the discourse, or discourses, of Quality and Safety, which had become cognate with 1000 Lives.

The discourse of Finance was reinforced and introduced from within the organizations by the CEs early in the reconfiguration. It was a consistent discourse across all the boards from a powerful subject position. The rhetoric of the discourse was, in part, that of break from the legacy practices and discourses of financial support from government. Historically government had “bailed out” health care providers when they overspent on budget and allowed the organizations to carry forward their overspends. The discourse included terms such as “no bail out,” an imperative to “break even,” and a need to “accelerate the pace of change.”

For example, in November 2009 Penninsula LHB record that “. . . a collective effort was required on financial delivery for the benefit of patients and services . . .” and in February 2010 City LHB holds a special meeting on finance where concerns are so strong that board is unable to sign off the accounts. These special meetings, with boards being unable to sign off the accounts, occur across the seven LHBs. The concern with Finance is extended as a requirement for the public as noted by Western LHB in March 2010, the “. . . public needs to understand that NHS Wales will remain in balance . . .”

This discourse of finance continues over the course of the study and while there is some consistency to the message the discourse itself is highly variable as boards are regularly told that they are on target to break even, often with the comment that they are the only board in Wales to achieve this, followed by the reporting of a large deficit some months later, or even within the same meeting. For example, Riverside LHB predicts a break even in the February 2010 minutes followed by a predicted deficit in July 2010. In November 2010 the minutes of Penninsula LHB record that the board is predicting a breakeven and is the only one in Wales to do so followed, later in the same minutes, by a recorded deficit of £15.5m (USD 24.8m) and the comment that breakeven is achievable if steps are taken. The meaning seems to be that of a requirement to achieve changes that reduces pressure on the health related spend.

This discourse of Finance, in turn, was linked to a further political discourse at the national level of providing a “world class health service” (Designed for Life: A world class Health Service for Wales, 2005) and the continuing discourse of Quality and Safety. The discourse of Finance and the discourse of Quality and Safety have the potential to conflict. Concerns were raised by external stakeholders about the level of attention being given to safety matters by the board and the increasing focus on financial matters early in the reconfiguration. These concerns were also raised from within the board (e.g. by Riverside LHB, May 2010) including, in one board, by the director of finance.

Faced with these competing pressures the boards appear to adopt statements, or tropes, which are used by the CEs. These statements combine the competing discourses of Quality and Safety and Finance, and attempt to satisfy the stated requirements of both the organization and the interested parties in the external environment. The development of this combined discourse is shown in Table 2.

OPEN IN VIEWER

Table 2.

Discourses on Finance and the Emerging Discourse That Links Finance and Quality and Safety. Most of These Statements Are Recorded as per the Chief Executives and Chairs of the LHBs

LHB	Discursive extracts
Hillside LHB	. . . ensure the organization meets its financial targets and breaks even . . ./. . . no additional funding available from the Welsh Assembly Government . . . (January 2010)
	. . . improve quality and patient care, which will result in savings./. . . focus was on services and patients . . . (May 2010)
	Further financial recovery measures were being implemented, but that the challenge remained to bring the Health Board into balance by the end of the financial year. The Board recognized and was committed to the need to deliver within the budget and resources available. /. . . the message of the seriousness of this financial position . . ./NHS was facing a significant challenge . . ./. . . the Health Board faced a significant financial challenge and that it must break even at the end of the financial year as the Welsh Assembly Government was not in a position to provide any financial support. (October 2010)
	the Health Board was here to ensure the delivery of safe patient care . . ./. . . the Health Board was here to make changes . . ./. . . the difficult financial context, both locally and nationally . . . identify and implement different solutions to ensure the delivery of high quality care./. . . focus being on quality and getting things right first time . . . (November 2010)
Riverside LHB	. . . improving consistency and achieving better outcomes for patients . . . the financial position of other Health Boards in Wales was similar to that within . . ./. . . significant challenges . . ./ . . . the Board to ensure a balance between quality, performance and financial resources (July 2010)
Penninsula LHB	. . . deliver a balanced year end position . . . (October 2009)
	. . . confirmed its commitment to patient safety . . ./. . . still operating at the ‘top end’ of performance compared to the rest of Wales . . ./. . . the key issue was one of safety . . ./. . . it was feasible that a breakeven position could be achieved . . . achieve financial balance (January 2010)
	. . . use of agenda ‘item zero’ to highlight patient safety issues . . . (February 2010)
	. . . committed to improving the health of the population and improving the safety and standards of care . . ./. . . entering a period of unprecedented financial challenge and together we will have to make difficult decisions . . . (June 2010)
	. . . a period of extreme economic circumstances . . ./. . . statutory obligations to meet financial breakeven . . . [the organization is] . . . committed to quality and safety . . . [and] . . . the pace of change necessary in the current climate represents an unprecedented challenge . . . (November 2010)
City LHB	. . . advise against projecting a deficit. LHBs had to break even, year-on-year, and the incoming UHB Board would need a report at its first meeting . . ./To reinforce the importance the Board is attaching to the management of patient complaints, the Complaints Committee will in future be . . ./. . . external support from WAG was unlikely to be available in the current financial year./. . . has to do better and to get it right, first time, every time./. . . reminded the Forum that the UHB has a statutory duty to break even . . . made assurances that there would be no compromise on safety and quality, and observed that getting things right the first time would cost less but that the right model of services was also needed. Tough decisions would have to be made on how to stop the overspend and issues such as recruitment, overtime and the nurse bank would be discussed. (October 2010)
	. . . improvements in efficiency./. . . importance of identifying additional savings measures./. . . difficult decisions to ensure end of year financial balance . . ./. . . commended the reports to the Board (in accelerating the Board agenda, whilst maintaining a clear commitment to safety and quality of patient care.) (November 2010)
Valley LHB	. . . areas of risk related to financial health . . ./. . . the key area of risk related to financial health . . . the Health Board was assuming the need to plan for a very limited increase in funding for management . . ./. . . reduce current levels of expenditure by , , , Considerably more work was required to produce a balanced Financial Plan (November 2009)
	. . . this was the most financially challenging year the NHS had faced for at least two decades . . ./. . . area of risk related to financial health . . ./The revenue position continued to be the major concern . . ./The key area of concern remained the break even target . . . (July 2010)
Western LHB	. . . will remain in balance this year and that, financially . . . is broadly in line with its predictions i.e. be in a break-even position at year end, having delivered on what we said we would deliver on, with plans in place to overturn the deficit . . . (January 2010)
	. . . current services are unsustainable . . ./Discussions are ongoing with WAG to put an agreement in place for a 3 year settlement process. To address the £72m (USD c. $109) deficit would require a 12% cost improvement program on budget which has never previously been achieved within public sector organizations./. . . it will take three years to get the organization into a financially sustainable position in order to give the population the services it both needs and requires . . . (September 2010)
Midcountry LHB	. . . maintain financial balance . . . (March 2010)
	Board held detailed discussion regarding the organization’s financial position . . ./. . . receive a Quality Impact Assessment (August 2010)

This position acknowledges both internal and external subject positions and recognizes the worth of both bodies of knowledge. The distinct discourses of Finance and Quality and Safety began to merge and “. . . finance is balanced by quality . . .” (Hillside LHB, May 2010). In this merging discourse there is the “need” to “drive” patient safety as hard as finance and, “. . . quality and safety remained the boards main priority . . .” (Peninsula LHB, July 2010).

The need to maintain a discourse on finance, which incorporates quality and safety is recognized by the WHSSC, which has influence over both the CEs and boards exercised from a distance.

The WHSSC has a budget of c. £600m (USD $910m) and replaces the body Health Commission Wales. All LHB CEs sit on the board and it needs at least four CEs to be quorate. It is chaired by Public Health Wales, which does not have responsibility for service provider organizations, and vice chaired by an LHB member (both of these board members are independent members). The WHSSC is responsible for governance of the LHBs and has wide ranging powers to audit quality and safety.

In the April 2010 Minutes the WHSSC members “. . . discussed in some detail the Plan and in particular the financial challenges locally to LHBs.” At the same time patients, patient safety, and safeguarding were put above all other considerations. In September, 2010, the NHS Wales finances for the LHBs are considered and “. . . those Members present supported to the options to address the adverse variance a letter would be sent to all Chief Executives as soon as possible to SUPPORT the measures as outlined in the report.” (Bold and capitals as in the original.)

In addition, there is an attempt at a consistency of approach through the use of a consultancy third party as the LHBs were, “. . . working with McKinsey on a local and all-Wales basis to look further at the wider service and financial framework and develop plans for the next five years.” (City LHB, January 2010). Western LHB also refer to McKinsey as being involved in a 1000 Lives workshop in 2009.

Although the discourse of Finance became “balanced” by a discourse of Quality and Safety, this was separate and distinct from both 1000 Lives and Patient Safety. This balancing seems to emerge most strongly toward the latter end of 2010. For example, the following discussion is present from September 2010 to November 2010 in Penninsula LHB:

[T] his situation was very serious, given the Health Boards duty to break even at the end of the financial year, whilst also meeting quality and safety targets . . . [ Penninsula is not] in the business of balancing finances at all costs . . . providing safe and sustainable services remains the overriding priority . . . [and even though the LHB has a cumulative deficit of £16.7m (USD 26.7m)] . . . patient safety remains of paramount importance, above all other considerations, but . . . pace of change in delivering efficiencies must increase . . .

Similarly, during a special meeting on finance City LHB makes reference to a statement of intent to deliver safe, high quality and sustainable patient care and notes the need to, “. . . accelerate the board agenda whilst maintaining a clear commitment to safety and quality of patient care . . .” Earlier on, July 2010, the Director of Finance at Penninsula LHB highlights the cumulative overspend and “. . . stressed that quality and safety remained the boards main priority . . .” Even earlier indications of this emerging balance are present in May 2010 when in Hillside LHB a member of the board notes that key focus seems to be on financial matters rather than quality of care. The response from the Chair and the CE is that the key focus is on services and patients and “finance is balanced by quality.”

It seems that over the course of the study 1000 Lives and Patient Safety lost prominence on the agendas of the boards, and quality and safety concerns were delegated to the relevant committee (PSQC) within the committee structure of the LHBs. However, at the same time two competing discourses, Finance and Quality and Safety, were merged at the level of the board. CEs introduced the discourse of Finance in an attempt to deinstitutionalize existing practice of relying on bail outs from government. While they didn’t advocate standard cost cutting measures such as redundancy—indeed, compulsory redundancy was ruled out by government—they did advocate increasing financial diligence and a change in attitude. Interventions such as 1000 Lives were a useful assumed means to reduce cost by increasing quality and safety of provision. At the same time the existing organizational institution of Quality of Safety requires maintenance, but is potentially challenged by a focus on “efficiency.” In order to balance out these competing forces a new discourse develops that combines both financial considerations and quality and safety considerations. This discourse is inscribed in the texts, the minutes, and in other texts such as the minutes of the WHSSC which has a governance and audit role in respect of the LHBs.

This is in contrast to the situation that emerged with Nursing. Nursing sought to deinstitutionalize those beliefs, practices, and so on, that did not emanate from Nursing and to institutionalize beliefs, practices, and so on, that did emanate from Nursing. In the case of Finance and Quality and Safety there is a more of a merging of texts and discourses to create a new text with a new discourse that seeks to account of both the competing discourses and to institutionalize the beliefs and practices that will allow cost savings to be made without disrupting quality and safety. From the texts it seems that although the combined discourse was constructed there was little evidence of change as the discourse around Finance continued to be erratic (e.g., with boards unable to clarify their financial positions and continuing to have difficulty signing off their accounts).

Discussion

When 1000 Lives was first introduced it was, in terms of the Phillips et al. model, a discourse with a set of texts with the potential to constitute an institution. This discourse, of 1000 Lives, was framed by the institution of Patient Safety. Given the importance attributed to Patient Safety at board level (e.g., Monitor, 2010; National Patient Safety Agency, 2009), and to 1000 Lives in terms of political and organizational support in Wales, there was every reason to assume that 1000 Lives would become a dominant discourse at the level of the LHB boards.

Instead, despite the contextual support, 1000 Lives does not achieve an institutional position at the level of the board. It is contested by texts presented by Nursing and Chief Executives and by a parallel discourse of Quality and Safety which seems to be distinct from the Patient Safety institutional discourses and texts. 1000 Lives is delegated to a committee although the use of its parts as levers for change remain available to the board. The discourses of Quality and Safety and Finance become merged to form a single text which the boards re-present to themselves and to internal and external stakeholders. These interacting discourses form “discourses of organizing” (Keenoy, Mars, Oswick, & Grant, 2000) in which there is a constant realignment of discourses, which become framed, in part, by the institution of the board (Selsky, Spicer, & Teicher, 2003).

1000 Lives became an object, a social artefact (Maguire, 2004; Maguire & Hardy, 2009), that was used to achieve particular ends, including ends other than that for which it was intended. Over the course of the study the boards’ utilization of 1000 Lives develops and it becomes a means to achieve culture change, improve leadership among nurses, and, contrary to the earlier statement, becomes a “methodology” to “drive up,” for example, clinical quality and the Patient Experience.

Conclusion

Written texts are used in discursive activity both as objects and as a means and these uses are interchangeable depending on the context. Authors fashion texts as objects in one setting that they use as a means in another setting. These texts range from documents through to sets of statements and bounded vocabularies and other symbolic representations. As the boards develop and record their emerging discourse they construct and reconstruct the social reality of their organization. The organizations are composed of fragmented and competing discourses with no finality to meaning making (Bakhtin, 1986), and continual conflict among multiple meanings looking for interpretive control (Eisenberg & Goodall, 1993), which the board must [re]arrange into a coherent, unified, and meaningful discourse. The discourse/s are inscribed in texts, in this case the minutes of the board meetings, and these texts contribute to organizational institutionalism and deinstitutionalizm; the construction and destruction of socially shared realities that are self-regulating. This is further complicated by the recognition that the minutes are themselves an institution.