Tom W. Reader

High-reliability organisations such as hospitals and other healthcare organisations operate under a paradox of learning: there persists a culture of organisational silence that discourages staff from speaking up about potential threats to (patient) safety, but staff feedback has long been identified as a valuable source of information for organisational learning and harm prevention. How do we overcome this irony and help the organisation learn? The purpose of this paper is to leverage a source of data where employees voice complaints about their organisation without the threat of exposure: the employee review website Glassdoor. In this chapter, we analyse reviews from 531 employees of the National Health Service (NHS) in the period from June 2008 to May 2017. From these reviews, we identify 1743 issues that employees voiced about. We thematically analysed these issues using an abductive analytic framework and found that employees voiced largely about management problems, relationship problems, occupational well-being problems, and patient care problems. Using human factors theory, we theorise that staff complaints fill a blind spot that patient complaints do not, by virtue of being privy to “back-end” organisational structures and functions. From a sociological standpoint, we also theorise that challenges arise in healthcare due to a clash between strategic action and communicative action and consider how better dialogical communication can bridge the two. We conclude by discussing the implications of using non-reactive data sources to aid in organisational learning and propose humanities-driven recommendations that can be adopted by healthcare organisations to improve staff and patient outcomes.

Safety listening—responses to voice acts aimed at preventing harm—can avert organizational failures like airplane crashes. Research often focuses on attitudes and perceptions of listening using self‐report measures; consequently, little is known about how safety listening occurs behaviorally and influences safety outcomes in high‐risk situations. Using directed and summative content analysis, we analyzed 45 transcripts of flightdeck communication before crashes and near misses to develop a framework of safety listening behavior in risky contexts. We also used abductive top‐down theorizing to identify the processes through which such behaviors prevent harm. We propose that effective safety listening behaviors engage with voice through action and sensemaking, whereas ineffective listening behaviors dismiss or exhibit token engagement with speaking‐up. Our analysis illustrates that engaging with voice enables teams to develop shared and accurate situation awareness of emerging risks, thus potentially averting accidents. Our findings demonstrate the importance of a behavioral approach to safety listening, illustrating that assessing listener engagement with safety voice—rather than attitudes—can provide an accurate and practical explanation for how safety listening influences organizational safety outcomes.

In established democracies, the threat of regulatory capture—often implicated in major crises—is usually less about financial mechanisms like bribery and more about the subtle social processes of cultural capture. But how exactly is cultural capture defined, theorized, and assessed, and what are its underlying mechanisms, manifestations, and impact? This article presents a systematic review (n = 39) of cultural capture, identifying 7 descriptions and 10 mechanisms. We consolidate these into five underlying concepts (CHAIN): Closeness; Homogeneity; Avoidance; Identities; and Networks. We introduce a parsimonious definition of cultural capture: social and psychological processes that bias regulators' beliefs and behaviors, aligning them with those of the regulatees and marginalizing alternative viewpoints. A key contribution of this article is developing a set of 33 novel behavioral indicators to examine these five concepts empirically and clarify their relationship to regulatory capture. Finally, the review highlights theoretical and methodological issues to address for the field to advance.

Measuring organisational culture is important for detecting the values and practices that increase organisational risk (e.g., unethical conduct). Self-report methods (e.g., surveys) are mostly used to study culture: however, due to reporting biases and sampling limitations, and the rapid advance of digital data, researchers have proposed unobtrusive indicators of culture (UICs; e.g., drawn from social media, company reports, executive data) as a supplementary methodology for identifying organisations at risk of failure. A UIC is a single measure of organisational culture drawn from data collected without engaging employees, and research using UICs is in its nascent stage. Although various data sources have been established for studying culture unobtrusively, researchers have yet to explore the application of multiple UICs drawn from different data sources. To investigate this, we developed an experimental battery of 83 UICs drawn from seven data sources (e.g., company earnings calls, employee online reviews, executive data), applying diverse analyses (e.g., natural language processing, quantitative analysis of behavioural data) to measure eight dimensions of culture (e.g., governance, integrity). We then applied the battery to assess 312 large European companies. We found that the UICs could distinguish between companies and different industries, and one dimension (customer focus) was associated with an outcome variable commonly used in culture research (Return on Capital Employed). However, we were not able to establish a coherent set of statistically reliable dimensions due to the clustering of UICs by data source. This clustering likely occurred because data sources reflected the values and practices of different stakeholders (e.g., employees, managers), which underscores a conceptualisation of culture that is focused less on shared values across an institution, and more on the values, priorities, and practices experienced by different sub-groups. Future research could structure UICs according to data sources and apply UICs to examine the causes of organisational failure.

Safety reporting systems are widely used in healthcare to identify risks to patient safety. But, their effectiveness is undermined if staff do not notice or report incidents. Patients, however, might observe and report these overlooked incidents because they experience the consequences, are highly motivated, and independent of the organization. Online patient feedback may be especially valuable because it is a channel of reporting that allows patients to report without fear of consequence (e.g., anonymously). Harnessing this potential is challenging because online feedback is unstructured and lacks demonstrable validity and added value. Accordingly, we developed an automated language analysis method for measuring the likelihood of patient‐reported safety incidents in online patient feedback. Feedback from patients and families (n = 146,685, words = 22,191,427, years = 2013–2019) about acute NHS trusts (hospital conglomerates;n = 134) in England were analyzed. The automated measure had good precision (0.69) and excellent recall (0.98) in identifying incidents; was independent of staff‐reported incidents (r = −0.04 to 0.19); and was associated with hospital‐level mortality rates (z = 3.87;p< 0.001). The identified safety incidents were often reported as unnoticed (89%) or unresolved (21%), suggesting that patients use online platforms to give visibility to safety concerns they believe have been missed or ignored. Online stakeholder feedback is akin to a safety valve; being independent and unconstrained it provides an outlet for reporting safety issues that may have been unnoticed or unresolved within formal channels.