When toxic leadership becomes a patient safety risk

How healthcare institutions protect harmful authority and what it takes to challenge it

Patient safety can depend on whether someone with less authority feels able to interrupt someone with more. A nurse may notice a deterioration that has been missed. A pharmacist may question a prescription. A trainee may suspect that a senior colleague’s explanation does not fit the evidence. In each situation, knowledge is useful only if it can enter the conversation and influence what happens next. Intimidation makes that passage more difficult.

Kathleen Saxton’s article in The Guardian, published on 24 September 2026, examines why companies protect managers whose behaviour damages the people around them. Her central question carries particular force in healthcare. Hospitals and health services depend on people combining their observations, acknowledging uncertainty and correcting one another. A leader who makes those activities personally dangerous can compromise the conditions under which safe care is delivered. [1]

The existence of the problem is well documented. In the 2025 NHS Staff Survey, 9.11% of respondents reported experiencing harassment, bullying or abuse from managers during the preceding twelve months; 17.01% reported such experiences involving other colleagues. The categories can overlap and combine several kinds of behaviour. They are reports of experience, not findings from completed investigations. Even with those qualifications, the results make it difficult to treat harmful workplace conduct as an occasional anomaly. [2]

Nursing research reaches beyond a single national survey. Leodoro Labrague’s systematic review, published in 2024, identified 21 studies examining abusive supervision and outcomes for nurses and care. It documented a range of adverse associations involving staff wellbeing, working behaviour and nursing care, while calling for stronger research on patient safety outcomes. The evidence warrants action to protect staff without pretending that every link to patient injury has been established. [3]

The terms also need care. A rude remark, sustained bullying, discriminatory treatment and retaliation against someone who reports a concern are different forms of conduct. They may coexist, but their evidence and consequences should be examined separately. Similarly, an urgent instruction during a clinical emergency is not equivalent to humiliation. A rigorous organisation must be capable of distinguishing necessary authority from its misuse without allowing urgency to become a permanent excuse.

What makes healthcare especially vulnerable is the combination of legitimate hierarchy and dependence. Clinical expertise matters. In an emergency, a team needs coordination and decisions. Yet the same professional hierarchy can make a junior person’s observation easier to dismiss. The problem arises when status becomes a substitute for examining the evidence, or when questioning a decision is interpreted as challenging the decision-maker’s standing.

Healthcare authority also extends beyond a formal management title. An influential consultant, educational supervisor or established member of a clinical team may shape access to training, opportunities and professional endorsement. A manager may control rotas, resources and performance assessments. From the perspective of somebody raising a concern, these forms of power can matter as much as the organisation chart. The relevant question is who can impose consequences, and whether their own conduct is subject to meaningful scrutiny.

That helps explain how harmful people can obtain positions of authority without needing to assume that every appointment was knowingly reckless. Organisations can mistake expertise in one domain for competence in another. Alan Benson, Danielle Li and Kelly Shue’s study of promotion decisions at 131 firms found that employers prioritised current sales performance over other observable characteristics that better predicted managerial performance. It was not a healthcare study and did not examine bullying. Its relevance is the question it poses for clinical organisations: what additional evidence is required before an excellent practitioner is given responsibility for other people’s development and working conditions? [4]

In healthcare, the argument for overlooking behaviour may be framed around indispensable expertise. A clinician brings referrals, performs difficult procedures or sustains a service that would struggle without them. A senior manager delivers targets under pressure. Those contributions can be valuable. But the judgement becomes incomplete if it excludes what happens to the people required to work around that individual: whether they ask for help, disclose mistakes or challenge a questionable decision.

The familiar distinction between technical brilliance and difficult behaviour can therefore be misleading. Clinical work is collective. One person’s contribution includes its effects on the information, cooperation and judgment of others. A technically accomplished practitioner who discourages questions may make a team less effective in ways that an individual performance record does not capture. A reasonable implication is that clinical competence should be assessed partly through its effects on the whole team.

Experimental research gives that concern a firmer basis. In a 2015 randomised study, Arieh Riskin and colleagues assigned 24 neonatal intensive care teams to receive rude or neutral comments from an observer during a simulation. Teams exposed to rudeness performed worse on diagnostic and procedural assessments. The researchers also examined information sharing and help seeking as pathways through which performance could be affected. [5]

This was a simulation, not a trial measuring injuries in actual patients, and the intervention was rudeness rather than sustained managerial bullying. Those limits should remain visible. Nevertheless, the finding challenges the assumption that disrespect is an unpleasant but clinically irrelevant feature of working life. Under controlled conditions, an interpersonal exposure altered how medical teams performed. [5]

A study of actual surgical care provides a different kind of evidence. William Cooper and colleagues examined 13,653 patients treated by 202 surgeons at two academic medical centres. Patients whose surgeons had accumulated more coworker reports of unprofessional behaviour were more likely to experience postoperative complications. The analysis considered patient and operative characteristics. As an observational study, it could not establish that the reported behaviour caused the complications, and unmeasured differences may have contributed. It did show why reports from colleagues deserve attention within patient safety work. [6]

Taken together, these studies make interpersonal conduct relevant to the assessment of clinical work. Staff deserve safe employment in their own right. The evidence about team performance and patient outcomes creates an additional responsibility: organisations need to examine whether protecting a powerful individual is making it harder for everybody else to deliver reliable care.

The Joint Commission recognised this connection in its 2008 alert on behaviours that undermine a culture of safety, updated in 2021. It identified intimidation and other unprofessional conduct as threats to communication, teamwork and care. The issue has therefore been recognised within formal patient safety guidance for many years. An institution cannot reasonably treat it as a newly discovered concern about office etiquette. [7]

Official inquiries show how the problem can persist despite warnings. The independent Paterson inquiry, published in 2020, examined the failures surrounding surgeon Ian Paterson across NHS and independent healthcare. Its findings extended well beyond an individual clinician’s wrongdoing to failures of oversight and response. The report recorded that healthcare professionals who raised concerns described bullying or aggression and a lack of support. The inquiry illustrates how reporting a concern about care could carry personal costs for those trying to protect patients. [8]

The East Kent maternity investigation, led by Bill Kirkup and published in 2022 as Reading the signals, examined a different setting. It documented dysfunctional relationships and evidence of bullying across professional and managerial groups. It also found that the trust had prioritised reputation management over openness with families, regulators and others. The report locates failure in the interaction between harmful conduct, divided teams and organisational defensiveness. [9]

The two inquiries concern different services and events, but both illustrate mechanisms described by Saxton: knowledge that does not become action, people considered difficult to challenge, and organisational responses that protect established authority. In healthcare, the consequences extend to families seeking explanations as well as employees trying to raise concerns.

There are two distinct failures here. Sometimes people do not speak because they fear the consequences. Sometimes they speak repeatedly and the institution does not respond adequately. Conflating the two allows organisations to prescribe more confidence training for staff when the unresolved problem lies in how management receives and investigates information. A reporting route has limited value if the person using it cannot discover what happened to the concern.

Dependence complicates the moral language surrounding silence. Someone who needs a training assessment, a secure income or a professional reference may understand the risks of challenging authority very clearly. Treating that person’s hesitation as a failure of courage neglects the choices available to those with greater power. Responsibility for creating reliable protection should increase with authority, rather than being passed down to the employee with most to lose.

Organisational conditions also matter. Justin Aunger and colleagues’ 2024 realist review drew on 148 sources to examine unprofessional behaviour in acute healthcare. It identified workplace disempowerment, organisational confusion and stress, weak social cohesion and cultures that permit harmful behaviour as important contributors. It also highlighted the greater exposure of staff from minoritised backgrounds. Its account broadens the explanation beyond difficult personalities and asks how the working environment enables particular conduct. [10]

An adequate response must therefore address working conditions alongside individual accountability. Replacing one manager may be necessary, yet the same pressures and permissions can remain for their successor. Conversely, workload should never become an all-purpose explanation that prevents an organisation investigating someone’s conduct.

The healthcare context introduces a further tension. Commitment to patients can make employees reluctant to leave a troubled service. Their sense of obligation may keep them working through conditions they would otherwise reject. It is worth asking whether an institution has come to depend on that commitment instead of addressing the conditions themselves. Staff dedication should not become evidence that the organisation can safely postpone action.

Psychological safety offers a more useful standard than an atmosphere in which everybody appears agreeable. In this context, it means being able to ask a question, acknowledge a mistake or raise a concern without expecting interpersonal punishment. Healthcare requires difficult conversations and the correction of poor practice. The aim is to make those conversations possible, including when the person who needs correcting is senior.

Ingrid Nembhard and Amy Edmondson studied 23 neonatal intensive care units and found support for a relationship between inclusive leadership, psychological safety and engagement in improvement work. Their study also examined how professional status shaped people’s experience of safety. The implication is practical: leaders’ responses to contributions matter. Inviting a question and then humiliating the person who asks it teaches a stronger lesson than the invitation itself. [11]

Changing that experience requires more than a workshop. Róisín O’Donovan and Eilish McAuliffe’s systematic review of fourteen interventions found mixed results for improving psychological safety, speaking up and voice. It noted limitations in outcome measurement and in what education alone could achieve. Training may help, but its completion is not evidence that a team has become safe to challenge. Organisations need to examine what changes in everyday interactions and in the handling of concerns. [12]

A reasonable starting point is to make behaviour part of appointment and promotion decisions. Clinical achievements and operational results should be considered alongside evidence about how a candidate supervises others, handles disagreement and responds to mistakes. Feedback needs to include people whose work depends on that person. Otherwise, selection will continue to privilege the impression made on those above while overlooking the experience of those below.

Once appointed, leaders should remain accountable for those behaviours. Repeated complaints, staff departures and concerns about supervision should be examined together, with appropriate confidentiality and a fair opportunity to respond. A pattern should prompt investigation, not an automatic finding of guilt. Equally, the absence of a formally upheld complaint should not be mistaken for evidence that no problem exists when staff describe credible reasons for avoiding the process.

Investigation arrangements need sufficient independence to examine powerful individuals and the actions of the organisation itself. That means considering conflicts of interest, explaining who can make decisions and protecting access to an appropriate alternative route when the usual chain of command is implicated. Protection also needs follow-up. It is inadequate to resolve the original concern while ignoring subsequent exclusion, lost opportunities or other credible signs of retaliation.

Fairness is essential to this approach. A conduct process must not become a convenient way to silence an unpopular clinician who raises legitimate safety concerns or challenges a cherished practice. Complaints should be assessed on specific behaviour and evidence, with proportionate responses and consistent standards. A culture of accountability must permit disagreement with management as well as protect employees from abuse by it.

Boards should ask questions that expose the difference between policy and experience. What happens after someone challenges a senior decision? Are repeat concerns connected? Does an employee who raises a problem remain able to progress professionally? Are patients and families heard when their account is uncomfortable? These questions are more informative than the mere existence of a policy or the number of people who have attended a course.

The point of examining toxic leadership in healthcare is to improve the reliability of care and the conditions under which people provide it. The evidence does not support reducing every failure to a bully, nor does it support treating conduct as peripheral to clinical work. It shows why institutions need to understand the relationships through which knowledge is shared, withheld, accepted or dismissed.

Saxton’s question about the protection of harmful leaders therefore becomes a question about healthcare’s ability to correct itself. A service can possess skilled staff and sophisticated procedures while making it difficult for inconvenient information to reach those able to act. The test of its leadership is whether a concern changes what happens next, especially when taking it seriously requires challenging someone the institution has learned to protect.

References

1. Saxton, K. (2026). “Toxic bosses: why do companies protect managers who make working life hell?” The Guardian, 24 September.

2. NHS Staff Survey Coordination Centre (2026). 2025 NHS Staff Survey: National Results Briefing. Prepared on behalf of NHS England. See the section “We are safe and healthy”, page 27.

3. Labrague, L. J. (2024). “Abusive Supervision and Its Relationship With Nursing Workforce and Patient Safety Outcomes: A Systematic Review.” Western Journal of Nursing Research, 46(1), 52–63.

4. Benson, A., Li, D., and Shue, K. (2019). “Promotions and the Peter Principle.” The Quarterly Journal of Economics, 134(4), 2085–2134.

5. Riskin, A., Erez, A., Foulk, T. A., et al. (2015). “The Impact of Rudeness on Medical Team Performance: A Randomized Trial.” Pediatrics, 136(3), 487–495.

6. Cooper, W. O., Spain, D. A., Guillamondegui, O., et al. (2019). “Association of Coworker Reports About Unprofessional Behavior by Surgeons With Surgical Complications in Their Patients.” JAMA Surgery, 154(9), 828–834.

7. The Joint Commission (2008; updated June 2021). “Behaviors that undermine a culture of safety.” Sentinel Event Alert, issue 40.

8. Independent Inquiry chaired by the Rt Revd Graham James (2020). Report of the Independent Inquiry into the Issues Raised by Paterson. Department of Health and Social Care, HC 31. See the Chair’s statement and the discussion of raising concerns, pp. 137–138.

9. Kirkup, B. (2022). Reading the signals: Maternity and neonatal services in East Kent — the Report of the Independent Investigation. Department of Health and Social Care, HC 681. See Chapter 1, key action area 4, and Chapter 4, paragraphs 4.83–4.101.

10. Aunger, J. A., Abrams, R., Westbrook, J. I., et al. (2024). “Why do acute healthcare staff behave unprofessionally towards each other and how can these behaviours be reduced? A realist review.” Health and Social Care Delivery Research, 12(25), 1–195.

11. Nembhard, I. M., and Edmondson, A. C. (2006). “Making it safe: The effects of leader inclusiveness and professional status on psychological safety and improvement efforts in health care teams.” Journal of Organizational Behavior, 27(7), 941–966.

12. O’Donovan, R., and McAuliffe, E. (2020). “A systematic review exploring the content and outcomes of interventions to improve psychological safety, speaking up and voice behaviour.” BMC Health Services Research, 20, article 101.