Learning is necessary to improve patient safety and quality in healthcare. Complying with the orthodox safety legacy (1), organizations have focused on learning from unsuccessful performances, such as accidents, incidents, or near-misses (2). However, contemporary safety thinking emphasizes the importance of learning from everyday work that goes well (3).

Healthcare organizations are often characterized as complex adaptive systems (4–6), where risk and safety are emergent rather than resultant outcomes, and where it therefore can be difficult unequivocally to link undesirable outcomes to specific causes. Healthcare system management is therefore gradually changing from a traditional focus on decreasing unwanted events and adverse outcomes to instead increasing opportunities for safe outcomes by enhancing the four systemic potentials for resilient performance (7).

It is now widely accepted that resilient performance requires the potential to respond, the potential to monitor, the potential to learn, and the potential to anticipate (8). This study focuses on how people can learn from everyday work that usually goes well and leads to acceptable outcomes, rather than learning derived only from work that does not go well and leads to unacceptable outcomes (9). Learning is often traditionally indirect and mediated by a team of experts that observe or analyze the work of others and provide delayed feedback—the lessons learned—to the others, usually in terms of what should not be done. But learning can also be immediate and allow those who work at the sharp end to learn directly from what they do. This form of learning can be done whilst working with others who share their insights about Work-as-Done—that is how work is actually performed in practice, as opposed to how it is imagined or prescribed. Although this type of cross-learning—defined as informal, or peer-to-peer learning that occurs through shared reflections and everyday attractions, is essential in most work settings, staff shortages, established orthodoxy and unpredictable demands often limit opportunities for team debriefings and reflections (10, 11). Teams rarely discuss what has happened during their working days unless a critical incident has occurred. Although learning from incidents is important, it is limited to what went wrong or did not work (also known as Safety-I) and fails to pay attention to what has gone well and what worked (also known as Safety-II) (12–14). Understanding how systems and processes normally operate not only focuses on what actually happens when work goes well but also provides more robust contextual knowledge that enhances how people make sense of events when they occur. Work that goes well represents “an ongoing condition in which problems are momentarily under control due to compensating changes.” (15).

For many years, it has been acknowledged that organizational learning does not happen in a single step but takes place through several stages (16). However, compared to the interest in organizational learning, relatively little attention has been paid to how people learn directly from their work, how this can change their underlying beliefs and espoused values [(17), p. 118], that determine how they do their work.

Direct learning occurs when individuals can discuss and reflect on their daily work, preferably when they still have good recall of it. By contrast, organizational learning is mediated in the sense that it is brought about by someone else, as shown in Figure 1. Cross-learning among clinical teams should be direct rather than mediated; people should learn directly from the situations they experience, rather than from how others analyze what was reported from their work. Learning should not be limited to what is conventionally reported, but reporting should be driven by the need of learning.

Direct learning has two other significant advantages over mediated learning. First, direct learning is immediate, with little or no delay, particularly once it has become a part of daily routines. Therefore, it can be directly applied in the context rather than later through recall or reconstruction. Mediated learning is inevitably delayed, either by weeks, months, or sometimes even years (Figure 2). Second, direct learning is specific and, therefore, relevant to Work-as-Done. Mediated learning is generalized to various degrees because it is interpreted by others and returned as advice or instructions, or even as formal institutionalized procedures (Work-as-Imagined).

The Resilient Performance Enhancement Toolkit (RPET) was developed to enable direct cross-learning among clinical teams by emphasizing daily reflective conversations, and common problem solving (18). RPET, engages teams in collaborative discussions guided by open-ended questions, such as “What went well today?” or “What challenges are we anticipating tomorrow?” This approach aligns with the principles of Safety II by focusing on understanding and reinforcing what went well, rather than just identifying errors, as in Safety-I.

The principle of RPET is to encourage people to consider issues, such as the following: •

What went well today?

This study reports the experience of a multi-site primary care system after the introduction of reflective daily learning sessions using RPET. Prior to the program, teams in the healthcare centers (HC) met regularly for projects and for steering the operation of the HCs, but there were no formal or informal meetings to reflect on the work done. Staff break times were often adjusted to ensure the continuity of operations which minimized opportunities for informal conversations among teams.

The aim of the study is both to describe the implementation of RPET across a healthcare system and to describe healthcare staff early experiences of using RPET.

The program was introduced in a phased approach across 27 HCs to facilitate cross-learning among teams and create opportunities to reflect on work. Initial training was provided to all 27 HCs; however, only 10 HCs proceeded to fully implement the tool as described in Table 1. In the table, an “X” denotes teams that successfully implemented the tool, while shaded cells represent those that did not proceed with implementation.

The use of the RPET varied among the 27 HCs. Seven of those HCs (consisting of 29 teams) that received training in Phase I consistently used the tool. Notably, despite the disruption caused by the COVID-19 pandemic, which significantly affected work processes, teams from ten HCs continued using the tool without any formal mandate.

A total of 72% of the clinical staff across all the HCs attended the virtual training in Phase II. At the time during the Covid-19 pandemic, many conferences and training programs were canceled, and a few were postponed and moved online, which might have contributed to the high rate of attendance for the RPET training.

In Phase III, the teams were invited to a one-hour discussion session to share their experiences using the tool. All 29 teams from the seven HCs participated in the consolidation session, shared their experiences, perceived benefits, and challenges. Interestingly, three of these seven health centers had reviewed less than 80% of their reported incidents, yet they demonstrated a strong commitment to learn from everyday work through the consistent use of RPET (Figure 4).

Moreover, the implementation of the RPET tool revealed several benefits and challenges expressed by the staff, which were compiled into a list as shown in Table 4. As these sessions took place during the pandemic, two out of the seven HCs used virtual platforms to share and discuss what was learned. The remaining HCs continued with in-person meetings, while observing infection control measures. The majority of teams, however, considered the program as an unnecessary activity and canceled it during the COVID-19 crisis.

Furthermore, the assessment of RPET meetings utilization across the seven HCs from 2020 to 2023 revealed significant variation in the frequency of meetings between the teams. In 2020, the frequency of RPET meetings ranged from daily sessions in some teams to irregular meetings held as infrequent as once per month. While some teams recorded up to 15–20 meetings per month, others met as little as two to three times a week or only once per month.

By 2021, this variation in frequency persisted, with some teams meeting almost daily while others maintained RPET meetings on a need-based schedule, meeting two to three times per month. The number of times RPET meetings were conducted varied from 256 times a year in one HC to none in others.

In 2022, teams maintained similar patterns, with several teams conducting RPET meetings either once per shift or 20–25 times per month. However, some teams reported conducting RPET meetings three to five times per month or only on an as-needed basis. The maximum recorded frequency for the year was 359 meetings in one team, indicating increased engagement.

By 2023, all teams were consistently maintaining RPET meetings, although the frequency of these meetings continued to vary significantly. Some teams conducted meetings 1–2 times per month, while others held daily sessions or met as frequently as 15–20 times per month. The number of learning meetings conducted ranged from a minimum of 24 times in one team to a maximum of 304 meetings in other teams in 2023. Throughout the four years, some HCs frequently shared feedback through an instant messaging tool at the end of shifts to ensure staff from other shifts could also benefit from the learning and coordination efforts.

Figure 5 below represents the frequency of RPET meetings held across different teams in the seven HCs from 2020 to 2023. The nursing team consistently reported the highest number of learning meetings while physicians and radiology maintained steady participating throughout the years.

The expert-led training sessions effectively increased the visibility of the team learning stories, the feedback shared by the teams during the consolidation sessions were analyzed using thematic analysis. The thematic analysis is summarized in Table 5. Staff across the seven HCs expressed their experiences through RPET meetings. Using the key themes identified in the analysis, their feedback was classified into four key areas of resilience: responding to challenges, monitoring situations, learning from experiences, and anticipating future issues (20).

Reaching safe outcomes for patients does not always happen in the same way or via expected paths but can sometimes be the result of daily and habitual adjustments (7). In this study, RPET meetings enabled teams to appreciate the value of sharing their experiences of the challenges they faced and how they routinely overcome these challenges. Learning moved gradually from being triggered by outcomes (incidents and near-misses) to being prompted by the overall resources and processes available that day, environmental challenges, and changes in patient influx. This is evident from the commitment demonstrated by the HCs, as noted in the results section, where even those that had not reviewed all their reported incidents consistently used RPET to address and learn from how everyday work challenges and opportunities were managed. This indicates that teams are committed to daily reflective meetings are not necessarily those who show consistent learning from incidents. Another key finding is the sustained increase in using RPET over time. This sustainability was likely driven by tailored training, one-to-one meeting sessions, and continuous feedback. Ros et al. highlighted that sustaining improvement in complex systems requires continuous engagement and the ability to adapt interventions to evolving needs (21). Consequently, this type of sustained engagement through RPET makes the learning approach particularly appealing to clinical teams, as it improves the staff's ability to notice and communicate what was happening in the meetings, thereby potentially improving the team's ability to monitor and anticipate challenges. However, Wahl et al. noted that although safety huddles can support system resilience through learning and responding, reflecting on positive events was often more challenging for staff compared to addressing adverse events. Improving processes and competencies will support the teams' abilities to respond to similar situations in future (22). RPET meetings facilitate learning by reflecting on how services are delivered despite challenges, aligning with Safety-II principles. As shown in Appendix A, Safety-II type questions prompt teams to reflect on successful adaptation, fostering a proactive learning environment and overcome challenges.

Learning is focused on “how” services are delivered despite challenges and has been shown to increase learning and responding capabilities in HCs conducting RPET meetings regularly. This is further supported by the patient safety culture survey results, which identified organizational learning and continuous improvement as key strengths, which reinforce the role of RPET meetings in sustaining a culture of continuous improvements (23). It was also reported that the teams felt that the individual experiences shared during reflection meetings provided collective learning. Peer learning, important for improving competency, was supported by the findings of Putri and Sumartini (24), who emphasized the value of shared experiences in clinical education.

While the initial intent of implementing RPET meetings was to provide opportunities for cross-learning among teams, the generated learning was found to improve four capabilities of resilient performance: monitoring, anticipating, responding, and learning. In addition to the cross-learning that resulted from the RPET meetings, the teams shared that they experienced more team engagement and cohesiveness. When urgent care nurses learned more about what triage nurses do daily, they gained more insight into why certain things happened in certain ways and avoided making assumptions or misinterpretations. When triage nurses learned more about patients' care and management post-triage, they understood how triaging patients is not only based on the triage policy (Work-as-Imagined) but also on the dynamic reality of that specific day (Work-as-Done). This communication in RPET meetings improved the understanding of the context, consequently reducing the tension between staff working in related clinical teams.

The results show that RPET meetings give teams opportunities to come together and share what otherwise usually goes unnoticed, namely, what works well. Initially, the staff were not sure how to talk about work when no incident occurred. The difficulty in detecting and noticing what goes well is acknowledged as a “real and serious problem.” (18, 22). Identifying 'safe adaptations' that result from the “fluency” staff gain in undertaking their tasks under varying difficulties, requires facilitated activities (25). When the teams started using Safety-II type questions (Appendix A), they realized that there were opportunities to learn from how challenges were overcome. Identifying Safety-II type question to facilitate Safety-II practices was demonstrated in other studies like Provan et al. (26) and Bentley et al. (27).

The study found that reflective daily learning sessions can significantly improve collective and cross-learning among clinical teams. By focusing on daily reflections, teams can identify opportunities for learning, uncover challenges, strengthen local adaptations, and improve communication. Face-to-face meetings helped to build trust among teams, impacting on information sharing and learning (15). The results of this study showed that providing different teams with the opportunity to meet daily for as little as 15 min has been shown to increase teams' understanding of environmental factors, organizational structures and dynamics, patients, communities, safety events, and employees' anticipation.

Organizations need to encourage local reflection practices as a key feature of learning, as mentioned by Edgar Schein: “a learning culture must value reflection and experimentation and must give its members the time and resources to do it” [(17), p. 119].

This study has several limitations that should be considered when interpreting the findings. First, the study was conducted within a single organization that is multi-sited in Qatar, which may limit the generalizability of results to other healthcare systems or cultural settings. Second, RPET adoption varied across HCs, with only a subset of HCs demonstrating sustained implementation during the study period. Thirdly, the data were primarily self-reported through facilitated discussions, which may introduce bias or overrepresentation of positive experiences. Lastly, the study did not include a comparator, limiting the ability to determine whether the observed outcomes were directly attributed to the RPET intervention. Despite these limitations, the insights generated offer valuable evidence on the feasibility and perceived value of embedding Safety-II principles through daily reflective practices in primary care settings.

This study demonstrated that it is feasible for healthcare staff to learn continuously from everyday work. Reflective meetings show how performance can be improved directly through experience at both the individual and team levels. This method eliminated learning delays.

The results of this study encourage further applications and practices of learning from everyday work, potentially transforming safety learning into encompassing learning from what has gone well in addition to learning from safety events.