Factors that drive team participation in surgical safety checks: a prospective study

The study setting was a 550 bed Australian tertiary centre, specializing in paediatrics, orthopaedics, vascular, cardiac, obstetrics/gynaecology, general, urology, ENT, plastics, neuro, ophthalmology, trauma, and thoracic surgeries. At the time of data collection, the perioperative department had 16/20 ORs commissioned and was staffed by over 300 nurses, physicians and ancillary staff. At this facility, approximately 20,000 surgical procedures are performed each year.

The selection of surgical procedures and participants was purposive and clinical participants were drawn from anaesthetics, surgery, and nursing, working across 12 surgical specialities. Healthcare consumers were selected based on their ability to reflect on their previous experiences as a surgical patient in the context of using checklists as a safety tool to prevent harm. Maximum variation in sampling [30] was used to achieve diversity in participants’ experiences and perceptions of team communications using the SSC, based on their professional and personal backgrounds.

Interviews were semi structured and were used to elicit clinician participants’ perspectives about the barriers and enablers of SSC implementation and use in practice. Interviews with the healthcare consumers were conducted to establish consumers’ expectations as recipients of surgical care. Semi structured interviews were digitally recorded and transcribed per verbatim. Prior to each interview, all participants completed a demographic questionnaire of characteristics (i.e., age, gender, education, clinical role) to contextualize the findings and describe the group. Participants were interviewed either individually or in groups, depending on their availability. Interviews were conducted at the participant’s convenience, away from the clinical environment. Group interviews were discipline-specific and were conducted with staff who belonged to the same staff category to diffuse potential status differentials [31]. Interview questions were based on Mitchie et al. [32]. ‘Theoretical Domains Framework’ (TDF) which targets aspects of individual behaviour relative to knowledge, skills, professional role/identity, self-efficacy, outcome expectations, motivation/goals, environmental context/resources, and social norms. For instance, to determine whether knowledge was a barrier or an enabler, we asked clinician participants, “What is your understanding of the purpose of the SSC?”, and, “Can you describe how you do the SSC?” Examples of questions asked of healthcare consumers included, “Do you believe that using a checklist can help reduce mistakes or errors?”, and, “Do you think that the checklist could improve communication among staff members in surgery?” The lead author performed all interviews and the number of interviews was determined by the point at which data saturation occurred.

Approvals were given by the hospital and university Human Research Ethics Committees. Participants were given an invitational letter detailing study information and were required to sign a consent form. They were advised of their right to confidentiality and anonymity and to withdraw from the study at any time.

From the textual analysis of field notes and interview transcripts, analysis occurred in an iterative manner to develop the subcategories and categories. The process of category development involved breaking down, comparing, and conceptualizing data to enable recognition of emerging patterns to identify major categories and their subcategories [33‐35]. The extracted categories and their corresponding explanatory data were cross-checked among researchers to ensure consensus. In data analysis, investigator triangulation was used as all members of our interdisciplinary research team contributed to the findings using varied perspectives and discipline-specific expertise [36]. In keeping with an integrated knowledge translation approach [37], we presented the findings to our participants to verify their authenticity and accuracy at each stage of the analysis. Data saturation at the point of analysis was evident when no new information or categories emerged.

The hallmarks of rigour concerning credibility, auditability, transferability and triangulation [30, 33, 36] have been considered in this study. Credibility was evident in relation to giving participants opportunity to corroborate the authenticity of thematic findings with their experiences. An audit trail, supported by memos, linked to pieces of data confirmed the categories and subcategories. Transferability of findings to other similar settings was achieved through using a wide variety of participants that included clinicians and health consumers. We used a triangulated approach on two levels: First, we used different data sources enabling cross checking, confirmation and completeness of the data [36]. Second, the interdisciplinary approach we used enabled a comprehensive examination of the data set, reducing the potential for bias that is inherent in any study.

The category, using the checklist encompassed behaviours based on checking processes that occurred during the three checklist phases. Interview participants described how using the checklist enabled them to identify deficits and/or discrepancies in important patient information, thus minimizing or preventing potential errors or mistakes. Having multiple opportunities to ‘check and recheck’ information was considered imperative to keeping the patient safe and these repeated checks created “layers of safety” (Int 28, Surgeon). Some participants however warned that using the surgical safety checklist as a ‘tick-n-flick’ exercise may have the unintended effect of leading to complacency because team members were merely ‘going through the motions’ without listening to, or thinking about the information being collected. Surgical team members believed that ‘modifying and adapting’ the checklist enabled them to tailor their care to the particular needs of the patient as well as reflect the nuances of different surgical specialities. However some participants believed that the ability to modify the checklist or checking process had to be carefully balanced with the need to standardize the items covered in the checks to ensure that important information was communicated and was not lost or omitted during handover exchanges. Another component to checklist use was about ‘being inclusive and patient-centred’—to ensure that the patient, where possible was given the opportunity to participate in the conversation around their details and to speak up when the information was inaccurate. From a patient’s perspective, healthcare consumer interviewees believed that the conversation with the patient around the checklist items minimized the “distance between you [as patient] and the healthcare professional” (Int 7, Healthcare Consumer).

The category, working independently illustrated team attributes that shaped the ways surgical teams coordinated checklist activities, and which was largely, defined by their roles, and the associated task-related workflows. The majority of clinical participants believed that physicians should ‘take ownership’ and ‘lead the process’ of checking, particularly during the time out phase. While they did not always lead timeout, physicians advocated strongly that they should coordinate the timeout check as they ultimately had responsibility for the patient. In relation to overall accountability, this statement was typical of sentiments held by surgeon participants, “I'm the one who takes the absolute risk” (Int 18, Surgeon). During the observational audit, anaesthetists and surgeons were observed to initiate 20 and 40 % of time out checks respectively. The influence of ‘being task focussed’, ‘working in isolation’ and ‘working in silos’ was obvious in the way team members emphasized discipline-specific information and the limited coordination in checks conducted as a collective. Throughout the audit period, we observed that anaesthetic and scrub/scout nurses appeared to be automatically allocated different sections of the checklist. For example, during sign in, anaesthetic and scrub/scout nurses were regularly observed performing and documenting (ticking) their checks independently of each other, although these checks were part of the first phase of checks. These observations were confirmed in the interview data. The notion of role demarcation was again reinforced in ‘being discipline-centric’ as team members checked off particular checklist items based on their different discipline orientations.

The category, communicating checks with others highlighted the communication strategies used during the checking process. Strategies such as ‘making sure, double checking’, and ‘verbalizing information’ enabled team members to develop a shared mental model, ensuring that everyone was on the same page. The various checks that occurred as part of the preoperative and SSC permitted team members to reconfirm information at different junctures to circumvent or avert potential errors, inconsistencies or omissions in relation to consent, procedure/laterality/proximity, imaging, or allergies. Likewise, the deliberate act of verbalizing information offered participants the opportunity to engage in a dialogue, especially during time out. Clinical participants advocated strongly that the time out check in particular, needed to be more “patient centric” and done while the patient was still awake and able to contribute to the conversation, rather than after anaesthetic induction. Nurse participants considered the patient ‘hand over’ process during sign in and patient transfer from the operating room to the post anaesthetic recovery unit (PACU), critical to providing seamless, safe care. However when the checklist was carried out as a “tick box” process, team members’ ability to ‘ask meaningful questions’ and elicit relevant information from patients and other team members was reduced.

During field work, we mapped team communications in relation to who was present and who participated in each phase of the checking process, and the proportion of checklist items that were verified and confirmed. Figure 1 illustrates the personnel involved in information exchanges at each juncture of the preoperative and surgical safety checklists. Throughout sign in and sign out phases, individual team members performed checks and communicated these as the need arose. Sometimes when the scrub and scout nurses were busy setting up for the next procedure, the anaesthetic nurse who had performed sign in would enter the OR to handover information to scrub/scout nurses about patient consent, procedure, and allergies. Anaesthetists were not directly involved in the sign in process, performed their own preoperative checks, asking patients information about medical history, allergies, and medications. Overall, total item completion of sign in items was 52.1 %. Anaesthetic nurses were primarily involved in sign in while only the scrub and scout nurses performed sign out. We did not observe participation by any other team members during sign out. Notably, sign in and sign out were the most vulnerable checklist phases because of the increased risk for information loss or omission.

Team participation was at its greatest during the ‘time out’ phase. That is, all team members were consistently present and simultaneously exchanged clinical information. During observations, timeout checklist items pertaining to patient identity and procedure were always stated and verified as a collective, while site/side was checked and confirmed in 60 % of cases. Confirmation of checklist items; critical and non-routine steps and patient-specific concerns was observed in 40 and 50 % of cases respectively. Significantly, verbal confirmation of equipment availability and sterility by the team was not observed.

Our findings highlight that team attributes reflect independence on an individual level versus interdependence on a team level. Professional independence requires that team members possess a tacit knowledge of regimens of care, therefore permitting members to perform their work tasks in a synchronized, unified manner [38]. Clearly being self-directed and working independently may improve teamwork. Yet, a significant limitation to professional independence is the siloing that comes as a result of a focus on professional role, which constrains opportunities for deliberate, open discussion among team members as a collective [39]. Study participants’ expectations that team members would naturally perform their sections of the checklist, with little, if any communication about how this would or should be done reinforced a task-oriented approach. In some situations, a task-orientated approach is required (e.g., patient intubation, micro-surgery) but without explicit communication about performing the checks, vital information may be missed and/or not passed on. Professional identity is accentuated by the “tribal affiliations” of members and the contrast between physicians’ and nurses’ traditional role responsibilities and teamwork behaviours in surgery [38, 40‐42]. The disciplinary focus of team members necessarily determines what information is communicated and when, and with whom it is communicated [21]. Participants in our study described the importance of having a designated leader for each phase of the checklist, and resolutely believed that physicians should lead time out. Findings from a recent realist synthesis of 35 studies found that ‘professional identity’ characterized checklist implementation, which is discipline-specific and more likely successful when physicians are actively leading and participating in the process [18]. Team attributes is the biggest factor in checklist participation and drives the communication strategies used when performing the phases of the checks.

Our observational and interview findings suggest that checklist communications especially during the sign in and sign out phases occurred with individuals serendipitously, rather than actively or deliberately. Minimal communication in surgery is perceived as acceptable in OR culture—ostensibly effective because it demonstrates clinical competence which is considered the epitome of safety [43, 44]. However, minimal or irregular team communications can also conceal inconsistencies or omissions in information [22, 25, 45]. Communicating task-related information also enables greater situation awareness [42], and thus allows team members to build shared mental models. From a safety perspective, checklists are regarded as a defence against adverse events [46], and when handover is incomplete, the risk of error increases. Similarly, when checks are performed as a tick box activity, the risk of failure to detect errors increases [47]. In our study, participants’ perceptions of risk seemed only apply to active failures and not latent conditions. Notably, physicians actively participated only in the timeout phase of the checklist as they believed that this was the phase where errors were most likely to be detected, a point highlighted in previous work [48, 49]. Nevertheless, as study participants identified, communications around the checks must be meaningful and the information sought, relevant and useful. That is, team members must ask pertinent questions to attain accurate clinical information that will be used to inform clinical decision making.

Our findings show that the timeout phase of the checklist was most valued by participants, and hence, had the highest level of team participation and completion rates. Most likely because timeout is the one time that the team whole is together with the patient. The sign out process was not a team-based activity, albeit that scrub and scout nurses performed checks at various time points during the procedure while multitasking as opposed to performing these checks during wound closure. Participants perceived sign out items as nebulous and unclear, and held disparate perceptions of when to initiate the sign out phase, likely because of conflicting individual workflows. Competing workflows consequently led to potential discrepancies or incomplete checks. That sign in and sign out checks were not confirmed and validated by the entire surgical team is largely driven by team attributes and communication strategies. Does it really matter that a team-based approach is not used for the sign in and sign out phases? Deliberate confirmation with oral validation of checklist items promotes closed loop communication and gives others the opportunity to ask questions and clarify concerns [50, 51]. Ideally involving the patient in confirmation and validation of information during the sign in and time out phases also adds another layer of safety. While errors in specimen labelling were not observed during fieldwork, several nurse participants reported mislabelling errors that were picked up during the sign out phase. Implementing behaviour change interventions aimed at educating and reminding staff about performing the safety checks, i.e., who does what, when and how may partially address gaps in participation. However, the broader issue around team attributes requires behaviour change interventions that address professional culture and systemic issues such as workflow.

Despite the use of a single hospital site; we spent extensive time interviewing participants and stakeholders including patients, thus allowing diverse perspectives. Such diversity may allow conceptual transference of findings to other similar settings. Twelve months prior to the commencement of this study, hospital staff had relocated to a larger facility, where the geographic layout was very different. This may have impacted on team communications because of the size of the facility. Finally, the dissimilar interview methods may have given rise to different group dynamics but similar issues were discussed and data saturation achieved.