In this study we explored motivations and barriers to speaking up perceived by doctors and nurses in oncology. To the best of our knowledge, this is the first study which assessed speaking up behaviors in the context of cancer care. Participants reported many situations in which they felt comfortable to voice their safety concerns towards co-workers and supervisors. However, episodes of remaining silent were a common experience among staff and were reported by all professions and hierarchical levels. The fundamental motivation for speaking up was to protect individual, identifiable patients from harm. While doctors and nurses felt strong obligation to prevent errors and safety violations reaching the patient, they were not engaged in voicing concerns beyond this immediacy. Preventing future harm in “statistical patients”, improving safety systems, or contributing to a learning organization were not mentioned as motives to voice concerns. This is exemplified by episodes in which observers could not stop the error or questionable behavior in due time. Only rarely had the observers taken opportunity to discuss the event later. Thus, actors in these episodes did not receive feedback on their performance. This is an important and novel finding. From a systems and learning perspective situations in which "statistical, future patient lives" are at stake and threatened by risky behaviours, may be even more relevant. Even if a clinical error cannot be prevented from reaching an individual patient, speaking up post-hoc would be important to avoid the error recurring and to open opportunities for learning and changing behaviour. Further research is needed into low-threshold interventions that would facilitate post-hoc speaking up behaviours, such as after action reviews or medication event huddles [27, 28].
While our study confirms many of the facilitating and impeding factors affecting voicing behavior reported in previous research, we also report some novel findings which may be characteristic for oncology and similar areas of clinical care. Staff reported various barriers for voicing concerns and weighted anticipated benefits against negative outcomes, e.g., damaged relationships or humiliation of a co-worker. In particular, many respondents reflected in detail on whether the level of risk for a patient “justifies” speaking up and its associated costs. Questions and concerns relating to medication safety were much less likely to remain unexpressed compared to violation of safety rules, issues related to hygiene and isolation. Also, staff was highly concerned with damaging (often long and intense) patient-provider relationships, trust of patients, and putting additional burden on the severely ill, and this was a strong motivation to withhold voice whenever patients or family were present. Whether this motivation is as strong outside oncology has yet to be confirmed. Obviously, there exists an imaginary threshold of potential harm (likelihood and severity) which enters the “internal calculus” whether and how to voice concerns. Of note, many individuals were well aware of the trade-offs they make. It is concerning that nurses and doctors believe that they are able – often within a heartbeat – to accurately estimate the risks associated with a specific behavior in a specific patient, e.g., not using gloves for lumbar puncture in a child. This “relative” interpretation of safety rules caused dissonance in some oncology nurses and resulted in feelings of resignation and futility often termed as “acquiescent silence” in the organizational silence literature . We suggest that nurses and residents need to be encouraged by unit leaders to defend these safety rules. Importantly, supervisor’s attitudes to silence, and thus the micro-climate, are a key predictor of employees silence behavior .
Nembhard and Edmondson recently introduced the concept of leadership inclusiveness, which describes “words and deeds exhibited by leaders that invite and appreciate others' contributions”, including speaking-up . Based on survey data obtained in neonatology they report that leader inclusiveness predicts staff psychological safety, an important antecedent of speaking-up. This stresses supervisors’ role in establishing and reinforcing a “culture of voice” . The unambiguous safeguarding by leaders is of significance if resident rotation and frequent co-operation with external specialities makes the predictability of consequences of speaking up difficult.
The relevance of fears relating to damaging social relationships and being labeled negatively is supported by previous research both inside and outside healthcare [9, 13, 32]. However, in our study anticipated negative effects on the relation between patient and the actor were even more critical barriers to voicing concerns. The presence of patients, relatives and other co-workers during the error or violation made the trade-off between speaking up and remaining silent exceptionally difficult. Participants often reported using questions, gestures and non-verbal signals to point the actor to the problem, but this communication was not always successful. More research is clearly needed to identify voicing strategies that would be acceptable to observers and actors in such situations. For example, teaching anaesthesia residents the two-challenge rule, a conversational technique that is assertive (advocacy) and collaborative (inquiry), increased frequency and effectiveness of challenges towards other physicians in operation room simulations . Such instructional interventions could be transferred to settings outside the operating room and seem particularly valuable for care teams involving different professions, specialities, and hierarchies, like oncology. Based on the findings of our study, we can clearly conceptualize scenarios with high and low difficulty of speaking up. These scenarios could be used to discuss and train speaking up behavior with oncology staff. They could also be embedded in a survey study as vignette case stories to estimate the relative, quantitative importance of different factors. More generally, the results of our study could be helpful to develop health care specific survey measures to assess trade-offs in voicing decisions.
Our study generally confirms existing models of employee voicing behaviors [4, 13] but also suggests some refinements and adjustments. In particular, we suggest some important contextual variables which are likely to have moderating and mediating effects in the health care setting (e.g., the presence of patients and family). The role and prevalence of perceived unpredictability of speaking up situations as a result from less known co-workers warrants further study. Our extended model of voice and silence can be used to generate hypotheses and quantitative testing.
The main limitation of our study is associated with the design: Our findings are based on clinicians’ reports of motivations and barriers. As such, they rely on participants’ introspection and are subject to various biases. For example, clinicians’ judgments of “low harm” associated with others’ behaviors may result from post-hoc rationalizations of their own failures to speak up rather than from concurrent risk assessments. Ethnographic observation would be an alternative to study speaking up behaviors in the field. However, the effects of contextual factors on HCPs’ voicing behaviours are not easily accessible and controllable in observational studies. As Tangirala argues, “silence” is a nonbehavior and as such difficult to observe, and even more difficult to interpret . While qualitative observation is a strong and valuable method to gain an in-depth understanding of HCPs’ real behaviours, it is limited to actual occasions and, by definition, cannot be extended to answer 'what if' questions. We propose future research to employ a triangulation of methods, including qualitative observation and quantitative surveys under experimental designs. Also, the extension of simulation studies to outside the operating room would be highly valuable.