Abstract
Changing organizations is difficult. In this article, we analyze how sensemaking that follows the initiation of change projects relies on the interplay of prospective and retrospective aspects, and we elucidate how organization members’ frames develop over time based on this interplay. Our data, 38 in-depth interviews with nursing and medical staff held at four different points in time, reveal how expectations impact the dynamics of meaning construction in change processes. Our findings demonstrate that the frames through which actors make sense of change initiatives develop continuously, although the expectations embedded in them are ‘sticky’ to some extent. The degree of ‘stickiness’ depends on expectations that are formed through initial prospective sensemaking, as these expectations influence actors’ tolerance regarding dissonant cues. Change initiatives fail when this tolerance becomes exhausted. Our study contributes to theory on sensemaking and change by elaborating on the undertheorized role of prospective sensemaking during change processes.
Keywords
Introduction
Managing change is an important but difficult task. This especially applies to projects that aim to deliver second-order change that attempts to alter norms, values and attributions (Weick and Quinn, 1999). Accomplishing such change not only has economic implications for many organizations, but also, for some, the consequences might literally be a matter of life or death (Grol et al., 2013). In the healthcare sector, organizations have increasingly adopted new standards and procedures with the aim of facilitating learning from errors and near misses and of implementing a safety culture (Tamuz and Thomas, 2006). However, developing a safety culture requires more than simply adding new formal procedures; it demands the fundamental change of an organization’s underlying meaning structures (Parker et al., 2006). Past change research indicates that sensemaking, that is, the social process of constructing meaning with the aim of understanding the nature of change (Weick 1995; Weick et al., 2005), and sensegiving, that is, the deliberate attempt to influence such sensemaking processes (Gioia and Chittipeddi, 1991; Maitlis and Lawrence, 2007), are important phenomena that shape how such a change in meaning structures unfolds (Gioia et al., 1994; Weick, 1995). Given the high failure rate of change processes (Beer and Nohria, 2000), it seems important to enhance our understanding of how the dynamics that underlie sensemaking processes contribute to the success or failure of a change initiative.
Current reviews of extant research on the topic indicate that more fine-grained analyses of the temporal orientation of the sensemaking process may improve our understanding of the concept (Brown, et al., 2014; Maitlis and Christianson, 2014; Sandberg and Tsoukas, 2015). Sensemaking can be conceptualized as a continuous process in which cues are interpreted through frames (Colville et al., 2012; Cornelissen et al., 2014; Weick, 1995). Analyzing how expectations become embedded in frames and how such frames develop make it possible to understand how attributions of meaning towards change initiatives develop over time. Whereas research has generally viewed sensemaking as a purely retrospective process, recent studies are increasingly emphasizing the need for a better understanding of prospective sensemaking dynamics (e.g. Rosness et al., 2016; Sandberg and Tsoukas, 2015; Stigliani and Ravasi, 2012). Investigating the dynamic processes of meaning (re-)construction in planned organizational change initiatives thus requires in-depth analyses of longitudinal process data.
In this article, we draw on such data to investigate the processes of sensegiving and sensemaking that precede and follow the implementation of a new error-reporting tool, CIRS (critical incident reporting system), which is introduced to facilitate learning from errors and near misses. In particular, our analysis focuses on the dynamics of initial prospective sensemaking as well as on subsequent meaning constructions relating to the change initiative through the continuous development of the frames. Our data comprise 38 in-depth interviews with nursing and medical staff held at four different points in time and enable us to examine how different expectations impact the dynamics of meaning construction in the course of a planned change initiative. We investigate how CIRS was implemented in two different hospital units – a trauma surgery unit and a gerontology unit – by applying a narrative approach (Pentland, 1999; Rhodes and Brown, 2005) in line with previous sensemaking research (e.g. Balogun et al., 2015; Brown, et al., 2008; Cunliffe and Coupland, 2012).
Our data from two units with different starting conditions allow us to identify two sets of meaning construction dynamics, both of which led to a rejection of the change initiative. By analyzing the dynamics of how the same change initiative failed in two distinct units, our findings contribute to sensemaking and change research in various ways. First, our findings demonstrate that the forming of initial expectations about a change initiative relies on the perceived degree of fit between perceived problem pressure and the change initiative and show how such expectations become embedded in frames through prospective sensemaking. Sensegiving is important in this phase because it aims to create perceptions of such fit. Second, we demonstrate how sensemaking following the initiation of change projects relies on the interplay of prospective and retrospective aspects, and we elucidate how this interplay influences how frames develop over time. Our study therefore connects to seminal works that investigate how the interaction between frames and cues stimulates or hinders organizational learning (Colville et al., 2014; Cornelissen et al., 2014). In particular, our findings indicate that frames develop continuously, although the expectations embedded in them are ‘sticky’ to some extent. Their ‘stickiness’ depends on the initial forming of expectations as actors continuously compare past expectations and current experience to make sense of ongoing processes. A discrepancy between past expectations and current experience leads to perceptions of ambiguity. The tolerance of such ambiguity impacts how expectations about the change initiative develop over time. The perception of a change initiative becomes negative as soon as the tolerance of ambiguity between prior expectations and current experience is exhausted. Hence, our study helps to explain the dynamics of change processes and to understand change trajectories during planned change initiatives. Third, we elaborate on how initially formed expectations affect the implementation of change initiatives. In particular, our study enables us to develop assumptions about how addressing the perceived fit between perceived problem pressure and the change initiative influences the success chances of change initiatives.
Theoretical background: Temporal orientation of sensemaking and change
The concept of sensemaking provides a means of analyzing the social processes of meaning construction (Brown et al., 2014; Maitlis and Christianson, 2014; Maitlis and Sonenshein, 2010; Sandberg and Tsoukas, 2015). Such processes of meaning construction occur particularly when organization members encounter ambiguity and seek to clarify ‘what is going on’ (Maitlis and Christianson, 2014: 58). Sensemaking is thus an essential feature of planned change processes as organization members attribute meaning to change initiatives. For the purpose of our study, we draw on Gioia and Chittipeddi (1991: 442), who define sensemaking in the context of planned organizational change as ‘the meaning construction and reconstruction by the involved parties as they attempt to develop a meaningful framework for understanding the nature of the intended strategic change’. Such intended change usually also involves planned initiatives to offer alternative meaning constructions (Maitlis and Lawrence, 2007). This deliberate offering of meaning constructions is referred to as sensegiving. Sensegiving thus denotes the process by which an individual attempts ‘to influence the sensemaking and meaning construction of others toward a preferred redefinition of organizational reality’ (Gioia and Chittipeddi, 1991: 442).
Sensemaking of planned change as a continuous process
Current research emphasizes that organizational change is an ongoing process (Tsoukas and Chia, 2002). Studies that build on this ontological assumption increasingly apply a process lens (Langley et al., 2013) and investigate the dynamics of organizational becoming (Langley and Tsoukas, 2010). From this perspective, sensemaking is about punctuating the process of becoming, with pauses in order to understand what has been going on (Colville, 2008: 166). How actors then actually make sense of ongoing change depends on how they perceive it, as the process of sensemaking relies on the relationship of frames and cues (Colville et al., 2012; Cornelissen and Werner, 2014; Cornelissen et al., 2014; Maitlis and Sonenshein, 2010). According to Weick (1995: 111), ‘frames tend to be past moments of socialization and cues tend to be present moments of experience’. Meaning is created through constructing a relation between these two moments. Frames pursue experience into memory (Bruner, 1990, cited in Colville et al., 2012: 7). They therefore ‘serve as the retention system for images of past organizational learning’ (Colville et al., 2014: 219) and ‘help individuals to comprehend and predict the behaviour of others through stereotypical inference’ (Cornelissen et al., 2014: 701). Hence, frames guide attention towards cues that fit them. Furthermore, they serve as interpretation filters as actors make sense of present cues through frames shaped by past experience (Weick, 1995).
Extant research has differentiated between several triggers for sensemaking (Maitlis and Christianson, 2014; Sandberg and Tsoukas, 2015). Unplanned events, such as external shocks, that trigger sensemaking come unexpectedly and require fast reactions. Under such conditions sensemaking ‘occurs when discrepant cues interrupt individuals’ ongoing activity, and involves the retrospective development of plausible meanings that rationalize what people are doing’ (Maitlis and Sonenshein, 2010: 552). Studies investigating sensemaking after shocks or crises have advanced our understanding of how frames shape interpretations of reality (Brown and Jones, 2000; Colville et al., 2013). Making sense of planned change, however, requires dealing with the continuous interplay of experience and expectations over longer periods of time (Balogun et al., 2015; Sonenshein, 2010). Drawing on Weick (1995), Maitlis and Sonenshein (2010: 564) argue that ‘expectations connect with cues to create meanings. Individuals then filter subsequent cues against this meaning and gradually build up confidence about a definition of the situation’. During planned change initiatives, the definition of a situation may change continuously as individuals constantly generate new experiences, adapt their expectations about the change initiative, and alter how they make sense of it accordingly. In addition, sensegiving that accompanies change can be seen as an attempt to influence frames by providing cues and by shaping expectations about the meaning of that change (Gioia and Chittipeddi, 1991; Rouleau, 2005). How frames develop during planned change is important because the experience that shapes them influences the images that actors develop about the future. In this vein, Hernes et al. (2013: 2) reminded us ‘that causality between events is ascribed rather than inferred, and that it is the ascription of meaning upon past events that gives rise to a projection of future events’. Despite its practical relevance, extant research has tended to underemphasize how such ascriptions and projections are interrelated and how actors engage in prospective sensemaking during planned change (Brown et al., 2014: 272; Maitlis and Christianson, 2014: 94).
Prospective sensemaking and the formation of expectations
Prospective sensemaking can be defined as ‘the conscious and intentional consideration of the probable future impact of certain actions, and especially non-actions, on the meaning construction processes of themselves and others’ (Gioia et al., 1994: 378) or simply as ‘sensemaking processes where the attention and concern of people is primarily directed at events that may occur in the future’ (Rosness et al., 2016: 55). As retrospective sensemaking occurs when actors interrupt momentum through time to answer the question of what has been going on (Colville, 2008), prospective sensemaking refers to structuring the future ‘by imagining some desirable (albeit ill-defined) state’ (Gioia and Mehra, 1996: 1229; Stigliani and Ravasi, 2012: 1234). Because prospective sensemaking has long been viewed purely as a derivative of retrospective sensemaking (Sandberg and Tsoukas, 2015: 18), it has remained ‘underresearched and undertheorized’ (Stigliani and Ravasi, 2012: 1234). However, understanding prospective sensemaking is important because ‘a focus only on its retrospective aspects neglects the historical arc or temporal embeddedness of sensemaking’ (Maitlis and Christianson, 2014: 96) and because currently ‘the importance of anticipation is missing’ from sensemaking research (Sandberg and Tsoukas, 2015: S24).
Although recent works emphasize the need to better understand prospective sensemaking (Battles et al., 2006; Gephart et al., 2010; Gioia et al., 2002; MacKay, 2009; Stigliani and Ravasi, 2012; Wright, 2005), hitherto few authors have investigated prospective sensemaking empirically, and very little is known about prospective sensemaking in the context of planned change. Wiebe (2010) demonstrated that the way managers framed time influenced the way they perceived a change initiative. Gephart et al. (2010) analyzed how past and present sensemaking orientations are used to create future images, and found that actors drew on schemes for action developed in the past and projected those schemes into the future, thereby creating hypothetical entities. Stigliani and Ravasi (2012) analyzed how the interplay between conversational and material practices impacts collective meaning construction. They contributed to our understanding of prospective sensemaking with their development of a model that shows that prospective sensemaking consists of interrelated cycles of retrospection. Their study, however, remains relatively silent on the role of prospective sensemaking in the context of organizational change. In a recent investigation of prospective sensemaking in a healthcare context, Rosness et al. (2016) found that prospective sensemaking was a continuous process based on a capacity for anticipation, enabling surgical teams to collaborate smoothly. Although these studies deliver important insights into the future-oriented aspects of sensemaking, several questions with regard to prospective sensemaking dynamics remain open. For example, there is little knowledge on how sensegiving and the organizational context affect the development of initial expectations during change processes. In addition, the question remained unanswered as to how the interplay between experience and expectations unfolds over time as a change initiative is implemented and how this interplay affects how actors make sense of the change initiative. Although previous research on future-oriented sensemaking emphasizes the importance of the interplay between retrospective and prospective elements of sensemaking (Gephart et al., 2010; Stigliani and Ravasi, 2012), its effect on the trajectory of change processes is still under-researched.
Methods
To investigate the dynamics of sensemaking during the implementation of CIRS, we opted for a qualitative research approach (Silverman, 2016). Qualitative approaches make it possible to grasp the complexity of organizational change processes (Flick et al., 2004). They are especially advantageous when knowledge about organizational phenomena is unsatisfactory (Eisenhardt, 1989). Langley et al. (2013) emphasize that creating process stories is a particularly well-suited strategy for moving from description to explanation when analyzing case study data. Drawing on a narrative approach, where stories are treated as abstract models, thus enables researchers to build or refine theory.
Research setting
We investigated the trajectories of change processes by analyzing the implementation process of CIRS, which is an anonymous critical incident reporting tool designed to improve the safety culture in hospitals (Staender et al., 1997). The basic assumption behind CIRS is that by reporting and analyzing near misses, actual errors with fatal consequences can be prevented because latent errors can be addressed systematically (Reason, 1990). Even the mere practice of discussing errors can make a difference, as hitherto it has been – and partly still is – a taboo to talk about them in the field of medicine owing to the social norm of individual accountability (Carroll and Quijada, 2004) and traditionally little system-oriented thinking (Tucker and Edmondson, 2003). Breaking that taboo may itself contribute to establishing a safety culture.
A pilot project took place from 2006 to 2008 with the largest hospital operator in Vienna. Fifteen units implemented CIRS with the aim of learning from near misses. The implementation of CIRS was part of a more extensive risk management project, which also included a risk audit conducted by an external agency. During the risk audit, the agency evaluated risks and created unit-specific risk landscapes. Six months later, the units’ risk landscapes were reevaluated to investigate whether changes had taken place.
The process of the CIRS project should run as follows. Employees post their reports electronically. For each unit, there are two to three reporting circle (RC) managers, who have exclusive access to the original reports. They check them for adequacy and delete any information that could be traced back to the actual person. After that, they forward the reports to the analytics team (AT), which consists of four to five doctors and/or nurses from the unit. The AT discusses the reported near misses, develops plans to eliminate potential sources of failure, and informs the whole unit about these plans. The unit’s senior managers then prioritize the AT’s suggestions and support their realization.
Data collection
To understand sensemaking dynamics during the implementation of CIRS, we conducted interviews at four points in time. Two units were in focus: a trauma surgery unit and a gerontology unit. We opted to investigate those two units in depth because both units were facing different challenges resulting from the variation in the level of risk, that is, the probability and severity of adverse events that members of both wards face in their everyday work. Interviews were held with six physicians and four hospital nurses (three physicians and two nurses per unit) at each of the four intervals, amounting to a total of 38 interviews (two interview partners dropped out in the last round). The interviews were semi-structured and included open, narrative questions. The answers were digitally recorded and transcribed; each interview lasted approximately 40 minutes. The first interviews took place before the implementation of CIRS in order to identify current behavior and procedures concerning the handling of adverse events, errors and near misses. The same people were interviewed a few months after the implementation of CIRS once the employees had had the chance to get used to the tool. The third round of interviews was conducted about 2 months later in order to investigate how the RC managers and the AT were working. The final interviews were held about 1 year after implementation. Thus, the data made it possible to grasp the processual dynamics of sensemaking by investigating how expectations develop (Langley, 1999).
To investigate the impact of sensegiving activities, we analyzed documents (kick-off presentations, web pages, etc.) and included questions aimed at understanding the expectations created through sensegiving activities. Combining multiple data sources facilitated triangulation and helped to increase the validity of empirical findings (Jick, 1979; Patton, 2002).
Data analysis
Although the processes in both units displayed several similarities at first sight, comparing the dynamics of the development in both units later revealed some important differences. In order to analyze these differences and the change dynamics behind the sensemaking processes in both units, we opted to apply a narrative approach (Brown et al., 2008; Dunford and Jones, 2000; Sonenshein, 2010). Because we understand sensemaking as a social process (Maitlis and Christianson, 2014), we regarded language as the central facilitator of collective meaning construction. This is consistent with the basic idea underlying the narrative approach, which is that language is a constitutive and not just representative element of social reality (Balogun et al., 2015; Rhodes and Brown, 2005).
Applying a narrative approach is particularly useful for analyzing change processes because it requires researchers to order ‘sequences of actions and events in a chronological and generally logically consistent manner in ways that explain equivocal happenings and outcomes’ (Brown et al., 2008: 1039). In our conceptualization of narratives, we followed the ideas outlined by Pentland (1999: 711) and understand narratives as constructs that ‘help explain the relationship between events in a process’. Because we already knew the outcomes of both processes when we started our analysis, we focused our analysis on finding explanations for why both processes unfolded in the way they did.
In a first step, we analyzed the stories provided by our interview partners to develop a better understanding of the context and of the process. We used inductive coding to analyze the emerging core themes of the change process and to understand how our interview partners constructed meaning with regard to CIRS (Miles and Huberman, 1994). Analyzing stories helped to reconstruct the frames through which our interview partners interpreted the change process, because ‘[t]he typical form of framing experience is in narrative form … it lends itself to storying’ (Bruner, 1990: 56, cited in Colville et al., 2012: 7). The deliberate sensegiving activities undertaken by managers from the hospital operator and members of the quality management department can be understood as cues that were similar for members of both analyzed hospital wards. We could access any further cues that shaped the frames of our interview partners only indirectly by analyzing how the frames changed over time. We found that at every point in time the analyzed stories contained a retrospective aspect, focusing on previous experience on dealing with CIRS (‘what has been going on’), and a prospective aspect, focusing on expectations regarding the future development of CIRS (‘what will be going on’). The frames that guide how actors created meaning around CIRS hence relied on prior experience and also incorporated expectations regarding the change initiative.
In a second step, we constructed composite narratives from the central themes identified in the interview partners’ elaborations on how the change process progressed in different phases (Balogun et al., 2015). Composite narratives are constructed by researchers and serve the purpose of capturing collective constructions of meaning (Langley, 1999; Sonenshein, 2010: 483). They rely on bringing together multiple fragments of stories told by several actors and are closely connected to the concept of frames (Cornelissen, 2012). Quinn and Worline (2008: 505), for example, emphasize that ‘[n]arratives, like frames, are socially negotiated constructs that organize people’s thoughts and actions, but narratives do so in a temporal sequence that explains how particular orderings of actors and actions lead to particular social arrangements’. Hence, we understand composite narratives as expressions of the frames through which actors made sense of the change initiative. Focusing on composite narratives at the four points in time helped us to develop an understanding of why and how frames developed over time. Investigating the entire process allowed us to identify sequential patterns that help to explain the reasons for shifts in meaning attributed to CIRS over time by making visible the antecedents and consequences of actions and events (Pentland, 1999; Sonenshein, 2010). To find explanations for the trajectory of the change process, we centered our analysis on the interplay of expectations regarding CIRS’ usefulness as a tool for improving a safety culture, the perception of its actual usefulness, and any discrepancies that originated from the relation between these. During the research process, the co-authors of this article analyzed the data independently and met regularly to share their impressions and discuss their findings.
At T1 we focused on the development of initial expectations. In this first phase, we concentrated our analysis on the starting conditions in each ward and on the role of sensegiving. We identified perceived problem pressure as an important influence factor on the forming of initial expectations. We understood perceived problem pressure to be the perceived discrepancy between the current state and a desired future state. Because prospective sensemaking is about ‘forward-looking thinking to “structure the future by imagining some desirable (albeit ill-defined) state”’ (Stigliani and Ravasi, 2012: 1234), we found it important to analyze how actors imagined this desirable state in order to understand how they created expectations. Actors especially relied on subjective perceptions as to whether the change initiative was contributing to some desirable state. If actors perceived CIRS as a useful tool for alleviating the subjectively perceived problem pressure, we coded this as high fit between the change initiative and perceived problem pressure. If they did not perceive CIRS as a useful tool to alleviate problem pressure, we coded this as low fit between the change initiative and perceived problem pressure. We discovered that sensegiving activities played an important role as they aimed to develop the actors’ understanding of the change initiative in both units. In this way, initial expectations were developed based on actors’ perception of fit against their interpretation of sensegiving.
At the subsequent points in time (T2, T3, T4), we investigated the further development of frames. This depended on the continuous filtering of cues leading to an accumulation of experience with the change project, which was compared with established expectations and led to ongoing adaptations of expectations. The retrospective aspect of sensemaking comprises experience with CIRS (experience in tx), which was interpreted against the backdrop of previously formed expectations (expectations in tx–1). We found that a perceived discrepancy between previous expectations and current experience led to perceptions of ambiguity. If there was a high discrepancy between both, we coded that as high levels of ambiguity; if there was a low discrepancy, we coded that as low levels of ambiguity. The prospective aspect of sensemaking in each phase was displayed in expectations regarding the further utilization of CIRS in the ward (new expectations in tx). Ambiguity had an influence on the continuous forming of those expectations. Our further analysis then focused on this influence and the development of expectations, experience and ambiguity over time.
Findings: How prospective and retrospective aspects shape sensemaking during planned change processes
We present our empirical findings in three distinct steps to explain how expectations impact the trajectory of change processes. First, we describe the sensegiving activities undertaken by the hospital management. Second, we illustrate the trajectory of the CIRS implementation process in both wards. Third, we compare the findings from both wards in regard to initial expectations as well as the development of frames, and we provide alternative options for preventing failure.
Initial sensegiving
Before the start of the risk management project, which included CIRS as well as a risk audit, managers from the hospital operator and members of the quality management department held a presentation at both wards. They emphasized that the new system was meant to achieve the goal of improving patient safety. To secure a high level of commitment and ongoing support during the project, all participating parties – senior management, quality management and ward management – signed a project contract.
During their presentation, quality administrators focused on comparing a negative safety culture with the vision of a positive safety culture that could be achieved by implementing and utilizing CIRS. Managers and quality administrators explicitly stated that the aim of the project was to change the common mindset vis-a-vis errors from ‘I made an error – no one can know about it’ to ‘Here is a lack of safety – everyone has to know about it’. The presentation stressed that CIRS reports serve as the basis for eradicating systemic errors and therefore the work processes, the organization and the safety culture would profit from introducing the tool.
In addition, they tried to dispel doubt by addressing fears and to overcome fears by providing information. It was made clear that only near misses should be reported, but not personal accusations and no actual errors with negative consequences for patients. Thus, CIRS facilitated learning from errors without negative consequences for either those reporting or committing the error. Quality administrators also emphasized that taking part in the pilot project as well as reporting was voluntary, anonymity was guaranteed, and free support would be provided. Furthermore, it was promised that each and every report would be processed and that management would ultimately decide on the solutions to be implemented.
Finally, quality administrators drew on evidence emphasizing the positive effects of CIRS that were found on other wards. In order to illustrate the whole reporting – anonymization – analysis – solutions – implementation cycle, quality management presented two best-practice cases where near misses were processed (the confusion of two drug packages, and the dispensing of drugs to someone with an allergy to them). Some initial results from the ongoing pilot projects and experiences from other wards were also outlined. More than 200 reports in 2 years and positive experiences with the external risk audit were presented as success factors.
Sensemaking processes
In this section, we characterize the sensemaking processes in the trauma surgery (S) unit and in the gerontology (G) unit. The case descriptions in T1 contain information regarding perceived problem pressure in each ward and initial expectations about CIRS. With T2 and T3 we describe how the meaning construction around CIRS develops over time. As there were no more expectations about CIRS by T4, we analyze how interview partners evaluated the process retrospectively in the last phase. We start every phase description by presenting the composite narrative we distilled from our data. These composite narratives summarize the frames through which experience was interpreted at the respective points in time of our study (Table 1, available online as supplementary material, provides an overview of the processes).
Sensemaking in the trauma surgery unit
In the trauma surgery unit, expectations regarding the potential impact of CIRS were high at the beginning of the project and guided how actors framed the project over time. Divergence between initial expectations and subsequent experience led to an increase in ambiguity over time. The change initiative failed as actors resolved ambiguity by radically adapting (lowering) their expectations as soon as they lost confidence that the change initiative would enable the unit to achieve the desired target state.
T1: Before implementation – ‘Hope for improvement’
Composite narrative T1: Learning from errors is not part of our culture. We either ignore them, or blame each other and do not discuss them in a way that would help us to improve safety. Now we hope that CIRS will help us with that. We are optimistic.
Perceived problem pressure
Physicians in the trauma surgery unit evaluated the prevailing safety climate and the collaboration as extremely poor. There was no awareness of errors, and scrutinizing events was regarded as annoying and the current handling of errors and adverse events as suboptimal. Near misses were seen as part of the normal state of affairs. Accusations, the personalization of failure, and individual punishment prevented individuals from accepting responsibility for their actions. As a result, there was almost no general awareness of errors and adverse events. One physician expressed that ‘one gets the impression that the same errors happen over and over again’ (S1M3). Nurses in the same unit described their climate in more positive terms, but still felt that the practices of learning from errors could be improved. Generally, interview partners described the climate as unpleasant, with poor teamwork, shifting responsibilities onto others and a lack of motivation, and so they perceived a high level of pressure to change the culture with regard to dealing with errors.
Initial expectations
The promises connected with CIRS fell on fertile ground, and doctors and nurses alike developed positive views that became embedded in the initial frames through which the project was assessed. Expectations were high and there was hope for improvement in the ward’s safety culture. As one physician put it: ‘I am happy about that (and) I do only see opportunities’ (S1M1). All interview partners were open to – even grateful for – the implementation of CIRS, and regarded it as an excellent opportunity to reduce errors and to facilitate learning at the organizational level. The physicians in particular welcomed the implementation of CIRS and expected that it would help them to talk about errors without accusing someone. Moreover, CIRS was regarded as a solution to the many pressing problems on the ward, and was expected to improve collaboration and communication.
T2: 3 months after implementation – ‘It will get better soon’
Composite narrative T2: CIRS has started; reports have been handed in. Rather few have participated to date and some concerns have been raised, but it will get better soon with the first observable outcomes.
Comparing initial expectations with first experiences
Three months after the implementation of CIRS, eight reports had been handed in, the RC had anonymized the reports, and the AT discussed them in two meetings. Changes had not yet been observed. There had not been any general sensitization towards errors, and some ward members soon developed more skeptical views when they realized the complexity of the problems they faced; others had even forgotten about the project. The first real-life experiences with CIRS drew criticism. For the first time, anonymity was regarded as a relevant topic. One physician even denoted CIRS as a fig leaf needed for legitimization purposes only. The risk audit by the external company was mainly seen as not very helpful, and was deemed to have produced few new or relevant insights.
Prevailing optimism
At T2, our interview partners explained that the project was going in the right direction but was still in its early stages. Those ward members who were familiar with CIRS were still motivated, and acknowledged that it would take more time to see results. They stressed that a critical factor in success and in getting others on board would be identifying noticeable changes resulting from CIRS. Overall, despite minor problems, a spirit of optimism prevailed. In general, CIRS was seen as ‘ok for the moment, as it still has to be processed and proposals will be developed and then people will deal with it’ (S2N2).
T3: 6 months after implementation – ‘It really will get better soon, but support is lacking’
Composite narrative T3: There are continuously new reports, but the feedback is still lacking. We lack time to analyze cases sufficiently and are no longer really sure about the anonymity of CIRS. But changes will happen.
Serious concerns
In T3, interview partners reported that they had observed a further increase in other ward members’ concerns and reservations about the system. The handling of errors had not yet been transformed. There were a number of unprocessed reports, owing to a lack of time and high general workloads. In addition, some ward members increasingly doubted the anonymity of the system and argued that ‘colleagues become increasingly suspicious, as they feel that CIRS is being abused in order to spy on each other … I hear that from various sides and this kills the project’ (S3M1).
Remaining hope
Our interview partners expected that the positive implications of CIRS would soon become visible. Six months after its implementation, the number of reports had increased remarkably and several reports on the same main topics had been submitted. Meetings in the AT had taken place; cases had been edited and forwarded to the ward management. In future, information summarizing the main results on one page would be circulated to all colleagues every quarter. There was still a strong belief that the whole project would run as soon as initial results could be observed: ‘I think the pace will be faster from now on, because more and more is prepared’ (S3M1).
T4: 12 months after implementation – ‘CIRS has failed. Nothing has changed; nothing will’
Composite narrative T4: CIRS has failed, we have no concrete results and if we propose changes we are left alone with the implementation. Nothing has changed, nothing will.
Lack of feedback and improvement
One year after implementation, criticisms still circled around the same topics. There were delays in analyzing the reports because scheduling meetings had proved to be difficult. Visible changes simply had not occurred. Overall, nothing new had been learned from CIRS; none of the discussed solutions had been implemented, and as a consequence CIRS had lost its credibility. Frustration levels were especially high because ward members felt that, after the whole laborious process of reporting, analyzing and suggesting solutions, no support and/or resources were provided to implement them. A cultural change had not happened either, as talking about errors had not been internalized. Overall, ward members had lost confidence that CIRS could solve the problems they wanted it to solve.
Disappointment and negative views
Perceptions of CIRS turned increasingly negative. Specific statements for future expectations were not made; the emphasis was rather more on justifications of why it was currently not being used. Once highly motivated nurses and doctors were increasingly frustrated, owing to the prevailing feeling that the whole project had been set up for purposes other than actually improving patient safety: ‘… we have the feeling that it is not there for us, to some extent for the patient, but it has other goals’ and ‘we are left out in the rain alone’ (S4M1). A nurse concluded that ‘on paper the whole project system is perfect, with these anonymous reports … but it needs a certain culture, a certain willingness to adopt it and this isn’t and wasn’t present’ (S4N1).
Sensemaking in the gerontology unit
In the gerontology unit, expectations regarding the potential impact of CIRS were low to moderate at the beginning of the project. Low initial expectations set the frame through which experience was interpreted; as a result, there were only low to moderate levels of ambiguity throughout the process. The change initiative failed, as actors could not see the project’s relevance and soon started to gather justification not to use it.
T1: Before implementation – ‘No problem with CIRS’
Composite narrative T1: Discussing errors is possible, but not all have realized that yet. Improving the system, however, can only be positive. Let’s see if it can help us, although we don’t expect too much.
Perceived problem pressure
Physicians as well as nurses emphasized that only very few errors happen on their ward. Our interview partners were proud that the safety climate had improved in recent years. The ward had a history where, under previous suppressive supervisors, the climate was poor and few were willing to show individual responsibility. Although that had changed, as errors could now be discussed openly and supervisors encouraged staff to do this, not all employees realized that discussing errors was not equal to punishment. The ward also had a long history of initiatives that aimed to improve safety, which had led to an increase in administrative tasks. Furthermore, the future of the ward was unclear, as it was supposed to be relocated, but information in this regard was sparse.
Initial expectations
Particularly the medical staff hoped that reports developed through CIRS might help to further increase transparency and improve safety by giving suggestions as to how to deal with potential errors more systematically in the future. Although they argued that they did not need the project, physicians acknowledged that ‘there are always processes that can be improved’ and that ‘one cannot possibly be against improvements’ (G1M1). Nurses emphasized that CIRS was interesting because talking about minor errors could be a way of further improving safety. Overall, however, our interview partners were somewhat cautious with their assessment of CIRS and its potential consequences, and the dominant perception of CIRS was that ‘I don’t know to what extent my colleagues will notice it or participate … for me CIRS is not a problem’ (G1N1). Some raised doubts regarding the anonymity and the further processing of reports, as well as the willingness of their colleagues to contribute to yet another working group.
T2: 3 months after implementation – ‘Nice idea that doesn’t hurt, but also doesn’t fit us’
Composite narrative T2: The basic idea behind CIRS is good. However, it does not fit our culture. We still discuss all errors immediately face-to-face. By doing so, nobody has to worry about a potential lack of anonymity in the CIRS system, either.
Comparing initial expectations with first experiences
Since the implementation of CIRS, only a very small number of reports had been written, and these were handed in by ward management rather than by front-line healthcare workers. As a result, CIRS seemed to have sunk into oblivion rather fast. Interview partners reported several reasons for this. There was no common understanding about what constituted an error and which situations – for example, communication problems – should be reported. Our interview partners also stressed that they still talked about errors immediately in person and did not really need an anonymous system for that, quite apart from their reservation about using computers. Scheduling meetings in the AT appeared to be difficult. Further, the fear that CIRS might not be completely anonymous had increased. The risk audit, where potential risks were assessed and analyzed, intensified the feeling that the external assessors and that those in charge of implementing the system had not really understood the challenges of the ward.
Little hope for increased use in the future
Some physicians still emphasized potential benefits of the system. They acknowledged, for instance, that CIRS might allow the ward to receive information they had not had until now from other hospital units. Interview partners argued that CIRS might also help to overcome hierarchies, as it provided a new communication structure. However, low expectations had generally been confirmed and, despite some positive remarks focusing on what CIRS could do, there was a common view that CIRS might be useful in principle, but was not suitable for practice on the ward. The general approach for dealing with errors had not changed. Errors were still talked about directly face-to-face, then fixed and quickly forgotten. Interview partners acknowledged that it would make sense to report to CIRS in parallel as well, but no one did this. General interest was lacking, partly owing to low commitment and partly because of the impending relocation of the ward.
T3: 6 months after implementation – ‘No one asked for it’
Composite narrative T3: We still do not use CIRS. Our expectations regarding positive outcomes of CIRS are low to nonexistent. Actually, we never really asked for this project. Rather, it was imposed on us. It cannot work because it is not suitable for a ward like ours.
Non-use and accumulation of justification for non-use
There were very few entries in CIRS. Overall the system was ‘barely accepted’ (G3M3) and anonymity was doubted. Topics were still discussed immediately as it took time for reports to be discussed. The main assumption was that even if there were more reports, ‘nothing would happen, no one does anything, no one cares, people say, they are not responsible for that’ (G3N2). Interestingly, the narrative as to why CIRS was not used started to shift slightly in T3. For the first time, people stated that CIRS had been set up as a top-down initiative by the board and that its implementation had been more the goal of supervisors than regular staff: ‘it was just an official order by the general management to implement CIRS’ (G3M1). As one nurse put it: ‘nobody asked us if we wanted it … so now we have it but we do not want it’ (G3N2).
No integration of CIRS in future routine
Nurses and doctors alike emphasized that no one used CIRS and that the system had already been forgotten. As one nurse clearly stated: ‘now it is too late’ (G3N). Future changes derived from CIRS were seen as highly unlikely. The relocation of the ward deterred participation, and some people stated that error analysis should be conducted in the new location rather than in the current one; the few optimistic accounts were given somewhat half-heartedly.
T4: 12 months after implementation – ‘CIRS is sleeping; let it sleep’
Composite narrative T4: The idea was nice, but it just did not work out. One could say the project fell asleep. We have no intentions of waking it up again. Incidentally, we also believe that the tool’s main purpose is not to inform but to monitor us.
Lack of traces
CIRS had not left any traces in the ward. No learning effects had been derived from it. Shortly after CIRS was presented, there was no longer any support for it. As a consequence, ward members quickly lost their motivation to utilize the system. It was not clear what the content of reports should be. In addition, anonymity was in doubt and there was a fear that everything was traceable. Finally, it was also stressed that the whole process of reporting, anonymizing, analyzing, and then providing and approving solutions took too much time. As a result, ward members assumed that CIRS only added more bureaucratic procedures without being particularly helpful.
‘Fulfilled expectations’
Overall, there were few entries and the system never started to run smoothly. Because expectations were set very low, there was also little disappointment. In fact, our interview partners emphasized that because the project did not leave any traces, it turned out as they had expected. One nurse stated that ‘there is no interest in CIRS, no one makes any entries. That is what I expected and that is what happened’ (G4N1). A physician argued: ‘Expectations have been met, as CIRS in this form is not implementable in the system’ (G4M1). The perception of their own safety climate, however, remained positive: ‘not directly connected with CIRS, but we are reminded to talk more about errors, also in the team’ (G4N2). With a few words nurses and physicians described CIRS as ‘food for thought’ (G4M1) and ‘a missed opportunity’ (L4M3).
Cross-case analysis: Expectations, experience and the trajectories of change
In the two case descriptions above, we have elucidated how the implementation of CIRS unfolded over a period of more than 1 year. Below, we first analyze how initial expectations develop. We then present the two patterns of failed change that underlie both change processes. Finally, we outline alternative options for preventing failure.
Prospective sensemaking and perceived fit between problem pressure and change initiative
Prospective sensemaking regarding the future states that follow the implementation of a change initiative is influenced by the degree of perceived fit between perceived problem pressure and the change initiative. Perceived fit refers to the perception that the change initiative serves as an appropriate means of alleviating perceived problem pressure. Sensegiving plays an important role here because by giving sense (providing cues), actors attempt to influence how perceptions of fit develop in the initial stage of change implementation. Perceptions of fit or misfit then lead to the formation of concrete expectations about the change initiative. These initial expectations become important elements of the initial frames through which the project and its progress are subsequently assessed.
In our cases, both hospital wards were confronted with the same initial cues in the form of sensegiving activities organized by their common hospital operator. In the trauma surgery unit, there was a high level of perceived problem pressure. CIRS was understood as a means of addressing and solving multiple, diverse problems within the ward, such as communication and collaboration problems. The perception of a high-level fit between problem pressure and the change initiative led to commitment and to the development of high expectations. In the gerontology unit, the same sensegiving activities were assessed against the backdrop of a lower level of perceived problem pressure. Here, ward members developed more neutral views towards CIRS. A lack of belief that the ward’s problems could be solved by implementing CIRS led to a comparatively low level of perceived fit of CIRS and, in turn, to developing low expectations (see Figure 1).
Development of initial expectations based on perceived fit between problem pressure and the planned change initiative.
Continuous interaction between expectations and experience and its impact on the development of ambiguity vis-a-vis the change initiative
The expectations that actors developed within their frames were continuously updated through the processing of cues, as members of both wards were constantly assessing the project and its progress. Because expectations are embedded in frames and frames are experience-based, the frames evolved continuously through comparison between current experience and expectations about the future that had been created in the past. A discrepancy between previous expectations about the change initiative (in tx–1) and current experience with it (in tx) led to perceptions of ambiguity. Actors can either tolerate such ambiguity and sustain high expectations despite negative experiences (if they remain confident that the change project will succeed), or they can react to perceived ambiguity by lowering their expectations (if they lose this confidence). The degree of ambiguity tolerance depends on the expectations embedded in the frame and influences the processing of cues as well as how such cues are used to develop future expectations (in tx).
We found high levels of ambiguity in the trauma surgery unit. High initial expectations, however, also led to high levels of ambiguity tolerance in the early phases of the project and prevented actors from lowering their expectations as the project progressed. Despite initial concerns regarding the anonymity and few observable outcomes, optimism prevailed as reports were handed in, anonymized and discussed, and observing noticeable changes was regarded as a matter of time. However, the continuous accumulation of negative experience led to a slight but gradual reduction in ambiguity tolerance over time. On the one hand, there were numerous unprocessed reports as the scheduling of meetings to analyze them proved to be difficult. On the other hand, it became apparent that recommendations derived from the already analyzed cases were not implemented, owing to a lack of support and resources. Consequently, the ward members could not observe any tangible outcomes, and ambiguity tolerance was exhausted in the last interview round. As ward members no longer saw the change initiative as a suitable means of achieving the goal of alleviating problem pressure, they developed increasingly negative views towards CIRS and lowered their expectations accordingly. In the gerontology unit, low initial expectations led to a situation where a lack of positive experience with the project matched expectations and actors did not perceive ambiguity. Right from the beginning, very few reports were handed in, and scheduling meetings for discussing them proved to be difficult. Furthermore, doubts regarding anonymity increased and the unit continued with their established process of discussing errors in person. An argument that contributed to the decline of the expectations was put forward explicitly later in the process: CIRS was considered a top-down initiative that had not been asked for and did not fit. Similar to past negative experience with quality management projects, CIRS was regarded as only leading to more bureaucratic tasks. A lack of perceived ambiguity as well as low levels of perceived fit between problem pressure and the change initiative also reduced further engagement with CIRS. Views on CIRS declined steadily, as it played no role in the unit (see Figure 2).
Development of expectations, experience and ambiguity.
The role of expectations for facilitating change: Decoupling/coupling of problem pressure and change initiative
Our analysis shows that CIRS failed in the trauma surgery unit and in the gerontology unit for different reasons. In both units, however, these reasons are rooted in the perception of fit or misfit between the project and the perceived problem pressure, which was also the foundation of prospective sensemaking. In the previous section, we analyzed the reasons for failure. In this section, we outline alternative options for preventing failure based on our analysis of the role of expectations about change processes.
In the trauma surgery unit, exaggerated expectations about the change initiative led to increased levels of ambiguity tolerance. Views towards CIRS became negative as soon as past disappointment outweighed positive expectations. Thus, the major challenge in the trauma surgery unit for making change work would be the decoupling of perceived problem pressure and the change initiative in the initial phase of the project. Facilitating such decoupling requires communicating that a change initiative cannot solve all the problems perceived by organization members, but can alleviate some. Hence, decoupling is effective because it decreases the perceived fit between the change initiative and the perceived problem pressure, setting perceptions at a more realistic level. Providing positive experiences may help to shed light on domains where the change initiative can actually contribute to solving perceived problems. In addition, providing such positive experiences can contribute to closing the gap between expectations and experience. Closing this gap prevents high levels of ambiguity from developing, that in the long run lead to failure once ambiguity tolerance is exhausted.
In the gerontology unit, members had low expectations regarding the usefulness of CIRS. As a result, actors did not actively engage in implementing CIRS. As experience matched low initial expectations, ward members never questioned their initial expectations. The challenge in the gerontology unit for making change work thus would be the coupling of perceived problem pressure and the change initiative in the early phase of the project. Facilitating such coupling requires communicating in a first step that the change initiative can solve at least some of the perceived problems. Here, coupling would be effective because it increases perceived fit between the change initiative and perceived problem pressure by creating concrete expectations that can be addressed by the change initiative. In a second step, creating positive experiences could lead to closing the gap between expectations and experience at a higher level.
Discussion and conclusions
Our study examines change dynamics over time by investigating the development of frames. It shows that the prospective as well as retrospective aspects of sensemaking interact in all phases of the sensemaking process and continuously influence how meaning is constructed and attributed to a change initiative. A perceived discrepancy between past expectations and current experience leads to ambiguity, which is tolerated as long as organization members remain confident that the change initiative has the potential to alleviate perceived problem pressure. Observing and reacting to gaps that occur between expectations and experience therefore represent a fundamental challenge for organizations.
Implications for theory
By providing an in-depth analysis of the interplay between sensegiving and prospective as well as retrospective sensemaking in two different healthcare units, our findings contribute to theorizing about sensemaking and change in several ways.
First, by analyzing how prospective sensemaking unfolds before the actual implementation of a new system, our study extends seminal research on prospective sensemaking (Gephart et al., 2010; Stigliani and Ravasi, 2012) and heeds the call to develop more nuanced views on the temporal orientation of sensemaking during change processes (Brown et al., 2014; Maitlis and Christianson, 2014; Sandberg and Tsoukas, 2015). Our analysis builds on the understanding of sensemaking as the creation of meaning through the continuous interaction between frames and cues (Colville et al., 2013; Weick, 1995), and contributes to the literature by elaborating on how actors develop and adapt their expectations, which are embedded in frames and influence how cues are interpreted. In particular, we find that the initial development of expectations about a particular change initiative relies on perceptions of fit or misfit between perceived problem pressure and the planned change initiative. Our analysis also elucidates the roles of sensegiving for the perception of fit and of prospective sensemaking for the perception of change. By contrasting how similar sensegiving activities (cues) aimed at influencing actors’ frames are processed in different contexts, our study heeds the call ‘to systematically compare sensemaking efforts triggered by similar events across categories’ (Sandberg and Tsoukas, 2015: S21). Whereas prior research has particularly underlined the importance of sensegiving for influencing perceptions of change (Gioia and Chittipeddi, 1991; Snell, 2002), our findings indicate that the varying perceptions of pressure to make the change intervention succeed had a stronger impact on the sensemaking processes in both units than the ‘standardized’ sensegiving activities. Although both wards were confronted with the same initial cues, actors in different wards framed the change initiative almost contrarily, leading to different change trajectories. By underlining the context-dependence of cue-processing activities, we also connect to research underlining the importance of customizing change initiatives to influence actors’ sensemaking of them (Stensaker and Falkenberg, 2007).
Second, we demonstrate how sensemaking that follows the initiation of change processes relies on the interplay of prospective and retrospective aspects. In addition, we elucidate how frames develop over time based on this interplay and how their development impacts the trajectory of change processes. We believe that our findings help to bridge two different views on the nature of frames with the aim of creating a more nuanced understanding of how change initiatives develop. Some works on the development of frames have focused on their reifying nature. Cornelissen et al. (2014: 728), for example, demonstrated how framing ‘may escalate from being a provisional interpretation to a collectively held belief’. As a consequence of such escalation, frames gradually grow less perception-based because the conception that guides perception becomes rigid over time (Colville et al., 2012: 9). This reification is also what Maitlis and Sonenshein (2010: 564) refer to when they argue that ‘[e]xpectations are sticky and this is where the danger lies – as individuals grasp tenaciously onto familiar meanings’. Other works highlight the relationship between frames and cues as an ongoing source of learning. In a seminal study on that issue, Colville et al. (2014: 228) found that it is possible to counter the natural tendency of frames ‘to become taken-for-granted and slip from conscious attention only to be revealed later, at times of crisis, as out-of-date assumptions … by holding frames and cues in a process of constant tension, creating learning moments’. Our findings contribute to this debate by analyzing the development of frames during the implementation of a concrete change initiative. In particular, we show that frames develop continuously although expectations are ‘sticky’ to some extent. The degree of ‘stickiness’ depends on the expectations formed through prospective sensemaking, as these define actors’ ambiguity tolerance regarding dissonant cues. Although experience and expectations continuously interact and this interaction affects the perception of change, expectations are radically altered as soon as ambiguity tolerance is exhausted and actors lose confidence that their desired future reality can be attained. By analyzing how experience and expectations impact the development of frames, our study connects to the growing stream of literature that examines these sensemaking dynamics (Colville et al., 2013, 2014; Cornelissen et al., 2014; Hahn et al., 2014; Holt and Cornelissen, 2014) and to research that investigates the role of expectations for sensemaking (Gephart et al., 2010; Rosness et al., 2016; Stigliani and Ravasi, 2012).
Third, by elaborating on how frames can be influenced through connecting or disconnecting perceived problem pressure with a change initiative, our findings also have general implications for explaining change dynamics (Balogun and Johnson, 2005; Gioia and Chittipeddi, 1991; Sonenshein, 2010). Whereas sensemaking research so far has focused on the effects of overly optimistic (Kayes, 2004) and overly pessimistic (Weick, 1988) expectations in dealing with disasters (see Maitlis and Sonenshein, 2010, for an overview), by elaborating on the development of expectations about planned change initiatives over time, our findings help to understand how these expectations can lead to unintended consequences. In our study, we not only elucidate two different patterns of failed change, but we also derive alternative options for preventing failure. These options rely on activities focused on influencing the development of initial expectations. Decoupling perceived problem pressure from the change initiative aims to avoid ‘overstretching’ frames by lowering potentially exaggerated expectations that the change initiative cannot fulfill. Coupling perceived problem pressure with the change initiative is directed at ‘stretching’ organization members’ frames for creating curiosity and openness to change. Thus, decoupling and coupling both focus on influencing organization members’ frames by stimulating the creation of perceptions of fit between problem pressure and the change initiative. Perceived fit plays a major role in the development of the expectations that subsequently serve as yardsticks against which any progress of the change initiative is assessed. Consequently, observing how such organization members react to the cues provided, and how their framing of change develops, provides a foundation for situational interventions on the part of organizational decision-makers throughout a change process.
Practical implications
Our findings also have several implications for practice. They highlight the importance of setting and communicating expectations during change processes. Our findings draw particular attention to the importance of the connection between perceived problem pressure and perceived fit when forming initial expectations. Depending on their particular combination, we can conceptualize four ideal types of how initial expectations about change emerge. First, if perceived problem pressure and perceived fit are high, actors believe that a change initiative can solve severe problems. High initial expectations and commitment would be a likely result, as was the case in the analyzed trauma surgery unit. Second, the combination of high levels of perceived problem pressure and low perceived fit is a potential source of low expectations and defensiveness, as actors regard it as unlikely that a change initiative will actually contribute to achieving a particular goal or target state. Third, low levels of perceived problem pressure and low perceived fit also lead to low expectations, which are the consequence of a lack of interest in the project. This was mainly the case in the gerontology unit we analyzed. Fourth, low levels of perceived problem pressure and high levels of fit occur whenever actors do not feel the urge to change, but are curious about how implementing a change initiative may contribute to a general improvement of the situation (see Figure 3). Identifying and managing the connection between perceived problem pressure and perceived fit requires managers to reflect on the context within which a change initiative is implemented and to carefully engage in sensegiving that supports perceived fit without detaching expectations from actually realizable results.
Development of expectations based on perceived fit and problem pressure.
Additionally, our findings underline the importance of customizing change (Stensaker and Falkenberg, 2007). In our case, we also have the specific situation that sensegiving activities overloaded the change in question with such positive connotations that objecting to the change in advance was almost impossible (Kousgaard et al., 2012: 199). As with other change interventions, developing individualized programs can contribute to reducing insecurity. In addition, establishing psychological safety (Edmondson et al., 2016) is necessary to realize the full potential of ex-post error discussions for learning (Dunn et al., 2016).
Limitations and directions for future research
The study reported here draws on a longitudinal research design to compare sensemaking in two different hospital units. Certain limitations should, however, be kept in mind when interpreting the findings. First, we analyzed the development of prospective and retrospective aspects of sensemaking by conducting interviews at different points in time. As a result, we can only compare different snapshots of expectations and experience. More granular investigations of sensemaking as a continuous process would require ethnographic approaches, which might also allow researchers to consider more emotional aspects in contrast to our rather cognitive focus (Maitlis et al., 2013). Second, previous research emphasizes the importance of differences between subcultures, for example, physicians and nurses, in healthcare organizations (Morgan and Ogbonna, 2008). Although we found no substantial differences regarding the perception of CIRS in our interviews among members from different subcultures, future research could investigate the trajectory of prospective sensemaking over time in cases where there is a significant difference in initial expectations between subgroups. Third, our study enables us to analyze two different patterns of failed change. Although we believe that both cases represent archetypical patterns of failed change, further patterns as well as alternative explanations for what we found certainly exist. Institutional theory (Dacin et al., 2002) and normalization process theory (May and Finch, 2009) are two conceptual lenses that offer alternative explanations for the effects we found. In addition, contrasting our findings with cases where very low perceived problem pressure was present at the beginning or with cases where change was successfully implemented could deliver further insights into the dynamics of sensemaking in change processes.
Supplemental Material
Supplemental material for Prospective sensemaking, frames and planned change interventions: A comparison of change trajectories in two hospital units
Supplemental material for Prospective sensemaking, frames and planned change interventions: A comparison of change trajectories in two hospital units by Stefan Konlechner, Markus Latzke, Wolfgang H Güttel and Elisabeth Höfferer in Human Relations
Footnotes
Acknowledgements
The authors wish to thank Associate Editor Tim Kuhn for his invaluable input and support as well as the three anonymous reviewers for the insightful developmental feedback they gave throughout the review process. The authors would also like to thank Leonhard Dobusch, Martin Friesl and the SCANCOR-Weatherhead scholars from fall term 2017 (led by Frank Dobbin), for helpful comments on earlier drafts of this article.
Funding
This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
