How to Break Medical Hierarchies to Foster Team Collaboration
The resident notices the prescription error but stays silent. The experienced nurse sees a worrying trend in the vitals but hesitates to "bother" the specialist. The respiratory therapist has a better ventilation strategy but defers to seniority.
Patient safety literature has a name for this: the authority gradient, the difference in perceived power between team members that determines whether a concern gets voiced. The goal is not to eliminate command structure, because during a code or a trauma clear leadership saves time and lives. The goal is to make sure the gradient never stops safety-critical information from reaching the person who needs it.
An authority gradient is the perceived power difference that decides whether junior staff voice concerns. Structured scripts like CUS, the Two-Challenge Rule and SBAR work because they remove the need for individual courage. None of them work without Just Culture and non-punitive reporting behind them.
What is an authority gradient?
An authority gradient is the perceived difference in power, status or expertise between people working together. A steep gradient means junior members are unlikely to challenge or question senior ones. A flat gradient means concerns get voiced regardless of rank.
The concept comes from aviation, where cockpit voice recordings from crashes repeatedly showed first officers hinting at problems rather than stating them plainly. Healthcare adopted the framing because the pattern is the same: the information needed to prevent harm is often already in the room, held by someone who doesn't feel able to say it.
How much does communication failure actually contribute to harm?
The Joint Commission's root cause data from 1995 to 2005 identified ineffective communication as a root cause in roughly two-thirds of reported sentinel events. In its 2014 data, poor communication was associated with 489 of 764 sentinel events, around 64%. Communication has consistently ranked among the top three root causes.
Worth being precise about: the figure describes communication failure broadly, not hierarchy or intimidation specifically. Hierarchy is one contributor among several, alongside handoff design, workload and documentation. Versions of this statistic that attribute it to intimidation alone, or round it up to 70%, are misstating the source.
What does the evidence on psychological safety actually say?
Amy Edmondson's research on hospital teams is the foundation of this field. Edmondson set out to test an obvious hypothesis: better teams should make fewer errors. The data said the opposite. Better-led teams reported more errors, not fewer.
The explanation was not that good teams were more dangerous. It was that good teams were willing to admit and discuss mistakes, while teams without that safety hid them. She named the difference psychological safety: the shared belief that a team is safe for interpersonal risk-taking.
That distinction matters practically. If you introduce psychological safety work on your unit and reported error rates go up, that is very likely the intervention working, not failing. Leaders who expect reported errors to fall immediately often conclude the opposite and abandon the effort.
Google's Project Aristotle later identified psychological safety as the strongest predictor of team effectiveness across its own teams, which is what brought the concept into mainstream management practice.
Why do steep hierarchies persist?
Not because anyone designed them that way, and not because senior clinicians are uniquely difficult. Four structural reasons:
- Legal liability sits with the attending. The person who carries responsibility for the decision has a rational interest in controlling it. Flattening communication without addressing who is accountable asks people to give up control while keeping the risk.
- Time pressure favours command. Deliberation is slower than instruction. Under load, teams default to the fastest structure available.
- Medical education is siloed. Nurses, physicians, pharmacists and therapists are largely trained separately and meet for the first time on a ward, without shared vocabulary or shared norms.
- Emergencies genuinely need clear leadership. The habits that work during a code carry over into situations that would benefit from discussion.
Recognising these reasons is what separates workable interventions from exhortation. "Be more collaborative" fails because it addresses none of them.
What structured tools actually help?
TeamSTEPPS, developed by the Department of Defense and AHRQ, is the most widely adopted framework in US healthcare. Its value is that it gives people scripts, which removes the need for individual courage in the moment.
CUS
A graded escalation script. Anyone on the team can say:
- "I am Concerned."
- "I am Uncomfortable."
- "This is a Safety issue."
The words are deliberately standardised so that using them signals a formal safety concern rather than a personal disagreement. Some implementations add a fourth step, "Stop," halting the activity pending reassessment.
The Two-Challenge Rule
If a team member raises a concern and it is not acknowledged, they state it a second time. If it is still unaddressed, they are expected to escalate to a supervisor or activate the chain of command. The rule works by making persistence an expectation rather than an act of insubordination.
SBAR
Situation, Background, Assessment, Recommendation. A structured handoff format that equalises communication by giving the more junior person a defined slot for a recommendation, which is the part most often omitted when someone feels outranked.
Surgical safety checklists
WHO-style pre-procedure briefings in which every team member states their name and role. The introductions are not a courtesy. Knowing a colleague's name measurably increases the likelihood of speaking up to them later in the procedure.
Graded assertiveness
The individual-level version of the same principle: escalate the directness of your language in steps, from question to statement to explicit safety language.
| Step | What you say | What it signals |
|---|---|---|
| 1. Question | "Have we considered…?" | Opens a discussion without challenge |
| 2. Statement | "I'm concerned that…" | Names a specific worry as your own |
| 3. CUS language | "This is a safety issue." | Formal safety concern, not disagreement |
| 4. Two-Challenge | Restate, then escalate | Protocol expectation, not insubordination |
| 5. Stop | "I need you to stop." | Halt pending reassessment |
What has to change at the institutional level?
Just Culture and non-punitive reporting. If speaking up or reporting an event can damage a career, no script will overcome that. Just Culture frameworks distinguish human error, at-risk behaviour and reckless conduct, and treat only the last punitively. Without this, psychological safety training asks staff to take a risk the institution hasn't agreed to absorb.
Interdisciplinary structure, not just goodwill. Daily huddles with rotating speaking order, multidisciplinary rounds where each discipline has a defined slot, simulation training with role reversal, and reverse mentoring pairing junior staff with senior leaders. These work because they build participation into the structure rather than relying on individuals to assert themselves.
What can you do on your next shift?
If you're senior: introduce yourself by first name. Ask "What am I missing?" and wait through the silence. Publicly credit someone junior who catches your error, since that single act does more than any policy. Rotate who leads rounds.
If you're junior: use graded assertiveness and escalate deliberately rather than hoping to be understood. Put concerns in writing when verbal channels feel unsafe, because a written record changes the dynamic. Frame around the patient, not the decision-maker. Rehearse difficult conversations with peers before you need them.
For everyone: learn names. Acknowledge expertise regardless of title, since "you know this patient best" is often literally true of the person with the least seniority in the room. Name good examples out loud when someone speaks up and is heard.
Key Takeaways
- An authority gradient is the perceived power difference that determines whether a concern gets voiced, adapted from aviation crew resource management.
- The goal is reducing barriers during clinical decision-making, not removing command structure during a code or trauma.
- Joint Commission data identified ineffective communication as a root cause in roughly two-thirds of sentinel events, describing communication failure broadly rather than hierarchy alone.
- Edmondson's foundational finding was that psychologically safe teams report more errors, not fewer. Rising reports indicate the intervention is working.
- Structured scripts work because they remove the need for individual courage: CUS, the Two-Challenge Rule, SBAR and surgical safety checklists.
- None of it holds without Just Culture and non-punitive reporting, because otherwise speaking up still carries career risk.
The barrier to speaking up is social. The friction afterwards doesn't have to be.
No tool makes a junior nurse feel able to challenge an attending. That is culture, and it is built by the things above. What communication tools can do is make sure that once someone decides to raise a concern, reaching the right person isn't the next obstacle.
Explore the Resources HubFAQs
What is an authority gradient in healthcare?
An authority gradient is the perceived difference in power or status between team members that affects whether junior staff will voice concerns to senior staff. A steep gradient suppresses speaking up; a flat one allows safety-critical information to surface regardless of rank. The concept was adapted from aviation crew resource management.
How much do communication failures contribute to patient harm?
Joint Commission root cause data from 1995 to 2005 identified ineffective communication as a root cause in roughly two-thirds of reported sentinel events, and it has consistently ranked among the top three causes. Note that this describes communication failure broadly, not hierarchy alone.
What is the CUS technique?
A TeamSTEPPS escalation script: I am Concerned, I am Uncomfortable, This is a Safety issue. The standardised wording signals a formal safety concern rather than personal disagreement, so the words do the work instead of the speaker's confidence.
What is the Two-Challenge Rule?
If a concern is raised and not acknowledged, the team member states it a second time, and if still unaddressed, escalates to a supervisor. It makes persistence a protocol expectation rather than an act of insubordination.
Does psychological safety reduce medical errors?
It changes error reporting first. Amy Edmondson's foundational research found better-led teams reported more errors, because they felt safe admitting them, not because they made more. Units introducing psychological safety work should expect reported errors to rise initially, which indicates the intervention is working.
Doesn't flattening hierarchy create chaos in emergencies?
No. The goal is reducing barriers to voicing concerns during clinical decision-making, not removing command structure during a code or trauma. Situational leadership with clear roles is compatible with a flat authority gradient.
Sources & References
▾
