Connective Routines: How Medical Professionals Work with Safety Checklists

“Designing” routines—a complex matter

Since routines entailmultipleactors inmultipleinteractions, Feldman and Rafaeli (2002) suggest that routines make connections between individuals. People connect through the performance of routines and then create shared understandings on two levels: on a micro level (1) shared understandings about a particularinstance of the routine being performed in its specific context, and on a macro level (2) these connections lead to shared understandings about the broader goals of the organization, backed with ideas about what is appropriate (Feldman & Rafaeli, 2002).

These shared understandings are vital for organizations. Although organizations aim to “plan” and “implement” routines—that is, make a new artefact routine behaviour—there is an essential element of interpretation in routinizing these artefacts. “Seemingly routine behaviour in organizations frequently involves human beings making interpretations regarding appropriate actions to be taken in a particular context” (Feldman & Rafaeli, 2002, p. 321).

Managers often introduce a bunch of artefacts to create a new routine. Sometimes they get the result they want, “but often they do not” (Pentland & Feldman, 2008, p. 235). The artefact thus reflects the managerial idea of the new routine, but the concrete expression of the artefact might differ from the written rule on paper. It is essential to recognize that the routine in practicethat emerges might differ from the routine in theory(see also D’Adderio, 2008).This resembles Brunsson and Jacobsson (2000, pp. 127128) distinction between “standardising practice” and “practising standards.”

Complying with standards

Although in routine theory the difference between the two routine dimensions, artefacts, and their interaction is widely recognized, it is attractive to focus on implementing standards. The implementation of new artefacts might be considered a rather “technical” matter. New artefacts such as checklists are seen as simple and cheap solutions to transform professional practice (see, for example, Treadwell, Lucas, & Tsou, 2014, claiming that surgical safety checklists “represent a relatively simple and promising strategy”).

Studies that adopt this instrumental approach often apply a certain kind of measure to scorethe extent to which employees complywith a new standard. These studies mostly rely on professionals’ self-registration data, reporting compliance rates even up to 99% or 100% (see, for example, Fourcade, Blache, Grenier, Bourgain, & Minvielle, 2011; Urbach, Govindarajan, Saskin, Wilton, & Baxter, 2014). It is increasingly acknowledged, however, that checklist compliance in a direct observation study is “a different reality”(Saturno et al., 2014, p. 289). Observational studies report compliance rates that hover around 30% (complete checklist compliance) to 55% (partial checklist compliance) (e.g., van Klei et al., 2012; Rydenfält et al., 2013). Nevertheless, even the few observational studies in the field merely measure whether there is rule compliant behaviour, without taking notice of the context with all its existing work patterns in which the checklist should be adopted.

Moreover, when there is a focus on complying with standards, connections and shared understandings between participants are presented as an outcome of the artefact. Bliss et al. (2012, p. 766, emphasis added) state that “a surgical checklist is an inexpensive tool that will facilitate effective communication and teamwork.” So by these insights, artefacts are seen as simpleand inexpensive tools that will generate connective actions and improve teamwork.

Resisting standards

We might use different insights, which stress social and cultural contexts in which new artefacts are introduced. The implementation of new artefacts might be seen as a social intervention that interferes with taken for granted ways of working (Evetts, 2011). This, in turn, might lead to acts of resistance. Standards are kept at a distance, manipulated, ritualized and/or counteracted.

It is important to note that standardization has always been a crucial aspect of professional work. To reduce complexity, various standards focusing on diverging facets of professional work acted as a form of occupational self-control (Freidson, 1974). To claim a jurisdiction, professionals set educational standards, ethical codes, and codes of conduct. Until recently, these standards thus mainly regulated professional work as case treatment (Noordegraaf, 2016). They did not really interfere with how professionals work together and how they jointly (re)configure professional work, set against new (performance) expectations and demands.

Formal standards, however, are quickly seen as the ultimate bureaucratic instrument, prescribing what to do when and in what ways (Berg, Horstman, Plass, & van Heusden, 2000). This is considered an assault on professional powers, for at least two reasons. First, the standardization of medical work interferes with professional judgment. Professional autonomy enables workers to assess and evaluate cases and conditions and to make judgements regarding advice, performance, and treatment (Evetts, 2002). Standards that prescribe how we should (re)configure professional work interferes with longstanding professional arrangements and is seen/felt as “intrusion” (Evetts 2011, Levay & Waks, 2009; see also Kirkpatrick & Noordegraaf, 2015). Second, standardization of professional work creates opportunities for external parties to exert control over professional work. Increasingly, standards are used by external parties as tools to hold them to account (e.g., Timmermans, 2005). Studies have therefore emphasized the lacking willingness of professionals to implement new standards that cause conflicts with professional values (Freidson, 1994).

What is more, new standards increasingly cross the borders of professional groups. Although there is a shift towards more multidisciplinary service delivery which calls for stronger connections both within and between disciplines, the socializationprocess through which professionals internalize “ways of working” and construct a professional identity takes place within professional segments (Bucher & Strauss, 1960).

The sociology of professions has emphasized how professions not only powerfully shape their members’ perceptions of themselves, but also of others outside their profession, and the “appropriate” patterns of behaviour in relation to others (e.g., Burford, 2012; Freidson, 1994). This creates an emphasis on the difficultiesin generating mutual understandings. It is argued that the different subcultures with their internalized norms, values, and diverging jargon, make the creation of mutual understandings crossing professional boundaries problematic (Abbott, 1988; Lingard et al., 2004). Consequently, a recent and growing body of literature has focused on what professionals do to maintainprofessional powers (e.g., Lozeau, Langley, & Denis, 2002; Waring, 2007). This explains acts of resistance.

Recent empirical findings move beyond notions of mere resistance but report how professionals adapt or capture reforms to strengthen their own position (Levay & Waks, 2009; Waring & Currie, 2009). These more nuanced analyses, however, predominantly depart from an individual standpoint—how individual doctors cope with reforms.

Surgical Safety Checklist

The World Health Organization (WHO) launched its “Safe Surgery Saves Lives” campaign in January 2007. The main goal of the campaign was to improve the safety of surgical care around the world, by decreasing unwanted variety in surgical care and improve teamwork within the OR (Haynes et al., 2009). One of the final outcomes of this program was the Surgical Safety Checklist.

The checklist was designed after extensive consultation of a team of surgeons and anaesthetists (WHO, 2009). Based on the WHO format, the hospital under study introduced the following series of checks that have to be performed at three strategic points in the process: (1) a team briefing at 8.00 a.m. in which each operative patient is briefly discussed by the complete operating team (surgeons, anaesthesiologists, and assistants) to review critical items such as patient identity, planned procedure, required materials, and known allergies, (2) a time-out just before incision, when again these items have to be checked, and (3) a sign-outwhere crucial items have to be checked before the patient leaves the OR, for example, if all gazes and needles have been removed, and where team members have to agree upon and register proceeding therapies.

Information about the new checklist was disseminated among the staff during several meetings and by e-mail. In addition, the checklist was made available in poster format in every operating theatre as well as electronically in the software system.

Research setting

This paper presents fieldwork conducted in 2015 and 2016. Over the course of fourteen months, the researcher undertook multiple field visits to the surgery department of a teaching hospital in the Netherlands. Approval for the study was obtained via the hospital board and the heads of the various departments. In order to formally gain access to the hospital under study, the researcher was appointed as “a research assistant” and to ensure the privacy of patients, had to sign a confidentiality agreement (CA). Before commencing the fieldwork, the researcher visited each key informant to inform and answer questions. The conversations all lasted between 40 and 60 minutes.

Data collection

A focused ethnographic methodology was adopted because it enables a close observation of the day-to-day activities of medical professionals. “Focused” in this approach refers to its problem orientation, as in case of FE the topics of inquiry are pre-selected (see, for example, Higginbottom, Pillay, & Boadu, 2013). Although the focus of this study thus was clearly demarcated in advance—Surgical Safety Checklist, artefacts, ideas, action, connections—this qualitative method using an inductive paradigm to gain in-depth understandings differs from deductive (observational) studies that might fail to capture an in-depth perspective. We especially related the checklists to day-to-day complexities and the real-life mechanisms that affect how and why checklists are used (or not).

Data collection involved observation of everyday work using a shadowing technique, in situ informal interviews as conversations (Mishler, 1996) arising during the course of observation, the collection of documents and attendance at meetings. As such, the data presented in this article are not gathered from formal interviews and stem mostly from participant observation and informal conversations.

There was a specific pre-selected focus on the checklist, but as this study’s aim was to find out as much as possible about the routines and interaction of workflows professionals are engaged in, we decided to observe full working days. To get access to the normally “closed world” of the surgical department, a shadowing technique made a valuable contribution (see, for example, McDonald, 2005). We started with purposive sampling, where we, together with a “gatekeeper,” identified key actors. Subsequently, snowballing led to the selection of other relevant actors that were willing to participate. Although we only shadowed the contact persons, during our days at the surgery department, we also interviewed many additional respondents, like full professors, division leaders, medical doctors in training, OR assistants, and nurse anaesthetists.

Observations were carried out over a period of fourteen months. Observation days lasted for approximately 10 hours, providing a total of about 140 hours of observation, carried out across the different physical areas of the department, including operating theatres, anaesthetic rooms, holding, recovery, training and meeting rooms, corridors, coffee room and changing room. Table 1 provides an anonymized overview of the data collection.

This method to collect data generated possibilities to gather information about the three different routine dimensions. Firstly, observations provided information about the actual performances by specific people, at specific times, in specific places (performative). What do they do, especially together, when they treat cases? These observations also gave insight in the specific physical setting, and the presence of written rules and procedures (artefacts), including the checklists as a physical or digital artefact. Lastly, conversations provided information about the ideas and (shared) understandings of the participants (ostensive). The conversations also helped to reflect on the gathered data with participants, and link the three routines dimensions in order to get a grip on how they are interrelated.

Data analysis

The ethnographic field notes were jotted down in a notebook and meticulously worked out in digital format after every episode of data collection. Both observation and conversation data were imported into Nvivo10 software for the purpose of thematic content analysis. We started off with an initial coding scheme (Figure 2) listing the two routine dimensions and artefacts. These codes are presented as three separate codes for analytical purposes, though we acknowledge that in practice they are constantly interacting and thus might be more “blurred.” Although we included artefacts in the analytic scheme, most emphasis in this paper was on ostensive/performative dynamics. For both these codes, we developed a subcode that reflects the connections; on the ostensive dimension, they refer to shared understandings (ideas), on the performative dimension they refer to actual interactions. The unit of observation was “action,” whilst the unit of analysis was patternsof action.

Informed by the sociology of professions literature, we also included a sub-code “professional power” at the ostensive dimension, to incorporate professionals’ ideas regarding the new standard. During the empirical process, we identified other relevant themes and inductively added thematic codes. “Team stability” for example, emerged as an important factor explaining differences in routines that were created. Further, practical circumstances strongly influenced the ostensive and performative dimensions and were added as a code.

The empirical findings will be presented beneath, revealing conditions under which connective routines can emerge. The results are structured along the themes that emerged and are specifically discussed in terms of ostensive/performative dynamics. The data was used to explore the processes of connective routines as social, situated and ongoing activities. Our aim was not to trace the implementation process in a longitudinal study but to get an in-depth understanding of how surgical teams jointly work with (or against) the checklist.

Figure 2.

Initial coding scheme.

Checklist use by general surgery teams

An explicit purpose of the checklist was to strengthen connections among team members. With more than twenty operating theatres, this large teaching hospital not only has a significant labour force but also a high staff turnover, since multiple employees come and go for educational purposes. As a consequence, the “teams” vary, and people operating together who have never met before is quite ordinary. In order to familiarize team members with one another, the SSC not only includes checks concerning patient identity and intervention but explicitly stipulates that team members introduce themselves before a case by writing their name and function on the whiteboard.

Performing the checklist, but not connective

The performance of the checklist requires interaction among the various team members. Every individual participating in the team must be able to communicate about the various items and “pause” the process in case the requirements of items have not yet been met. The findings show that for creating connective patterns of action, the hierarchical position matters.

The following observation note reflects the performance of a time out procedure in the general surgery department.

Varying ostensive aspects: Equal teams vs hierarchy

The working situation hindered the possibility to immediately ask for further clarification—why did the assistant mumble? A conversation later on, however, focuses on the experienced hierarchical relations. This OR assistant is finalizing her education, and because of her educational program, she has worked at various surgical departments to get acquainted with the diversity of surgical interventions. The observation note reflects the conversation we had about the performance of the time out in various contexts.

OR assistant: The performance of the time out differs widely. In some instances, it is just very quick and superficial, while in other cases, it is a rather extended procedure in which all items on the checklist are also written down.

I: How do these differences occur you think?

OR assistant: I think it has to do with how approachable the doctors are, and whether it’s a ‘real’ team. It has to do with the atmosphere, whether there is a pleasant and open atmosphere. Sometimes, you have the feeling that we are all equal, and then it [the checklist] goes smooth. Especially with the older doctors, you notice that it’s more hierarchical.

This conversation shows how individuals high in the hierarchy play a key role in the emergence of connections. Team members refer to a “pleasant atmosphere” and the existence of “a real team” as requisites for performing the checklist together. The surgeons—who often lead the checklist procedure—are indicated as the actors responsible for the atmosphere in the theatre. If other team members feel free to speak out, they are more inclined to participate in the team discussion and interrupt when necessary. However, if the surgeon explicitly presents him or herself as leader of the team and others do not feel that they are “all equal,” it becomes more difficult to cross these hierarchical borders. In these situations, there is attention for the checklist, for example by the assistant who mumbles and completes the items by herself, but not in a connective matter.

Checklist use by specialized surgical teams

Although most surgeries are performed in “variable” teams and an explicit purpose of the checklist was to create firm connections among team members, there are a few subspecialties where teams work together in more stable compositions. Subspecialties such as thorax surgery and vascular surgery are forms of very specialized work that require more stable teams. Anaesthetists that work in thorax surgery, for example, only work in these specialized areas. Therefore, fixed teams emerge in which surgeons, anaesthetists, nurse anaesthetists, OR assistants, and specialists that operate the cardiopulmonary bypass machine frequently work together. Because of these frequent encounters, these teams have the possibility to create shared understandings about what has to be done and what is appropriate.

Performances: Deciding on the spot

Observations were conducted at both the departments of vascular and thorax surgery to see how these specialized teams work with the SSC. The observation day at the thorax surgery starts at 8.00am in the operating theatre. The team immediately starts with the time out—the second part of the checklist. Since all team members only have responsibility for operations in this OR today, everyone is present in time. The thorax surgeon starts the time out and checks the patient’s identity, allergies, and prosthetic devices, and he performs the procedure entirely from memory.

After the time out, the surgeon leaves the OR to scrub, while the residents, nurse anaesthetists, and OR assistants prepare the patient for the surgery. Interestingly, when the patient is asleep, the team members are making ‘fun’ with each other while doing their jobs, for example, by squirting water from injection needles in each other’s ears. People not only know each other by name, but they also seem to get along quite well and work in a “relaxed” atmosphere.

A couple of minutes later, the surgeon is operating the first patient of the day. While he is working, the next patient is already discussed in an informal way. Statements like “what shall we do”, “you tell me!” and “we’ll get there” pass by. Though the “plan” for the next patient is discussed on the spot—comparable to what a briefing stipulates—this conversation is not systematic and moreover, required equipment is not yet resembled at the start at the day, as the checklist prescribes.

Two hours later, the operation is finished. While the assistants are cleaning up the OR, the surgeon asks: “Did we do the sign-out?” The other team members nod approvingly. “Oh, I missed that. That’s not quite right actually.” He replies. When the next patient on the table, the surgeon takes the lead in the time out again. Just like the case before, he checks the identity, intervention and prosthetic devices out of memory. The performance here deviates from the prescribed items on the checklists; some items are not covered, whilst others (prosthetic devices) are added to tailor the checklist to the needs of this specific context.

Ostensive aspects: What actually “is” the routine?

A few minutes later, when the patient is on the table, and the time out has just been performed, we start a conversation with the resident in thorax surgery to ask him about the briefing. The observation note covers the short conversation.

I: Do you also have a team briefing?

Resident in surgery: This was the briefing

I: No, this was the time out, the last check right before incision of the skin.

Resident in surgery: Oh, no. We don’t have a briefing then.

Some confusion occurred, since the resident was convinced that they do work with the checklist—they indeed “performed” some deliberation regarding the patient. However, it turned out that this team had altered the checklist through recurring performances in such a way, that it deviated from the artefact as such. In this case, strong connections among team members—that thus already exist—seem to encompass a lot of trust among team members in which people are inclined to rely on each other’s performances rather than explicitly checking the items with one another. Table 2 summarizes the findings covering the internal routine dynamics.

Practicalities and unexpected events that affect routine dynamics

Although our initial focus was on the internal routine dynamics, the interaction with existing work routines showed great influence on these dynamics.

The surgery department schedules the various operations in which mostly one surgeon is responsible for the surgeries in the operating theatre planned that day, for example, a range of hip fractures or colon carcinomas. The anesthesiologists on the other hand, are responsible for at least two of these operating theatres at the same time, which not only means that they have to monitor two patients at the same time, but also that they have to attend two briefings and time outs “at the same time.” The observation record shows how the organization of the time out in the care process leads to “basic” irritations, simply because people have to wait for one another.

This situation reflects the importance of the embeddedness of the new artefact within existing practices. The organization of the care process makes it difficult to create a new connective routine within this high-paced, demanding environment. The checklist aims to connect the different professional segments and improve their collaboration, the lack of “fit” with the existing workflows, however, not only hinders the creation of a connective routine, but the basic irritations tend to reinforce segmentation and thus stimulate the opposite effect.

Comparable practicalities emerged for example with “two-part surgeries,” where two different surgeons perform different parts of the intervention (for example, a breast mastectomy followed by reconstruction), but have to be present for the briefing and time out together. This routine required professionals to wait for another, interrupt and align their tasks, which proved time-consuming effort. In these instances, the checklist routine interfered with existing workflows, which means that professionals have to improvise and decide “on the spot.” This might imply that the action patterns that emerge deviate from the “rule” as inscribed in the artefact. Nonetheless, these instances of “noncompliance” might very well be best solutions for the situation at hand.

Acknowledgements

We thank all the respondents for their participation in this study. We also thank the anonymous reviewers and ISA RC52 for their constructive comments and suggestions.