Leading Through Pressure Without Losing the Human Side

The unanswered concern is an executive risk

A new handoff requirement looks sound in the boardroom. On a busy unit, however, it adds a step at the moment staff are managing admissions, medication schedules, and competing care needs. Someone raises the problem. A manager thanks them. Two weeks later, the process is unchanged, and no one can say who considered the concern. The organization has collected information without establishing what should happen next.

The problem is not necessarily an absence of goodwill. It is an incomplete decision process. Executives cannot personally investigate every operational concern, but they can define how important concerns acquire an owner, how urgent risks receive immediate attention, and how staff learn the outcome. A credible response may be a change, a bounded test, a request for more information, or a reasoned decision to retain the existing process.

For CEOs, COOs, CNOs, medical leaders, and people leaders, the useful question is whether a concern can be traced through that process. A listening session is an input. A meeting is a setting. Neither establishes that the risk was assessed, that responsibility was accepted, or that a decision reached the people affected. Leadership attention should extend beyond invitation to reliable follow-through.

What the research supports, and what it does not

Psychological safety concerns the interpersonal risk involved in asking a question, acknowledging uncertainty, or raising a concern. Edmondson’s foundational 1999 study linked team psychological safety with learning behavior in manufacturing teams. It provides an influential conceptual foundation, but it was not a hospital intervention trial. A later review by Edmondson and Bransby describes a broader literature connecting psychological safety with learning, performance, work experience, and leadership. These concepts help frame executive questions; they do not establish that a particular huddle will prevent harm. [18] [6]

The distinction matters in healthcare. Montgomery and colleagues’ 2025 review examined nine studies linking psychological safety with observable patient-safety outcomes or reporting behavior. Eight were conducted in the United States, and the participants were predominantly nurses. Findings were heterogeneous, and the authors did not find grounds for a clear general conclusion about the relationship. Reporting more concerns can mean improved openness, worsening conditions, or both. A dashboard should therefore distinguish reporting activity from independently assessed harm. [1]

Grailey and colleagues synthesized 62 papers from 19 countries and found substantial variation in psychological safety across settings and participants. This argues against treating an organization-wide average as a complete description of staff experience. A favorable result from one professional group or work location may coexist with silence elsewhere. The practical implication proposed here is to examine who participates, who is missing, and how local working conditions differ before choosing an intervention. [8]

Intervention evidence also requires restraint. O’Donovan and McAuliffe’s review of 14 interventions found mixed results and limitations in measurement. Jones and colleagues reviewed 34 studies of speaking-up interventions and emphasized the influence of hierarchy, relationships, and local culture. A training course or new reporting form should not be described as a proven solution to deeply embedded organizational behavior. Implementation needs attention to the response staff receive and the environment in which they work. [13] [10]

Accountability and staff voice belong together

Psychological safety does not remove professional standards or decision authority. It should make relevant information easier to surface while leaders continue to distinguish urgent hazards, operational constraints, competing preferences, and conduct concerns. Respectful challenge does not entitle every proposal to adoption. Equally, positional authority does not make an impractical process workable. The task is to make the decision more informed and its ownership clearer.

Nembhard and Edmondson’s study of 23 neonatal intensive care units associated leader inclusiveness with psychological safety and improvement engagement. Its survey design supports attention to leadership behavior, while limiting causal claims. O’Donovan and McAuliffe’s separate review of enabling factors identified influences across organizational, team, and individual levels, including support, learning orientation, hierarchy, and inclusiveness. These findings suggest that an invitation to speak should be accompanied by conduct that makes participation credible. [17] [11]

For an executive, that conduct can be concrete: describe what is open to change, explain what is constrained, invite the people who perform the work to identify practical problems, and name the person who will decide. When a proposal is declined, explain the reasoning without attacking the person who raised it. If a concern involves misconduct or retaliation, use the appropriate confidential process rather than resolving it in a public huddle.

Define the concern before choosing the response

Kane and colleagues’ review highlights the inconsistency with which speaking up has been defined and measured. Their synthesis connects raising a patient-safety concern with reaching a person who has the power to address it. That distinction is useful for management: a concern can be voiced repeatedly without reaching someone able to change staffing, technology, policy, or coordination across departments. [3]

Ask the person raising the issue to describe the work, the circumstance, and the possible consequence. Avoid requiring a fully developed solution as the price of being heard. A staff member may recognize that information is missing without knowing which team controls the relevant system. The initial response should establish whether immediate action is needed and who is equipped to investigate, rather than demanding that the reporter solve a cross-organizational problem alone.

Dixon-Woods and colleagues’ qualitative work examines how people decide what counts as a concern they can voice. Okuyama and colleagues’ literature review identifies influences that include professional relationships, anticipated responses, and whether speaking is expected to make a difference. These findings support looking beyond communication confidence alone. Leaders should examine which concerns are dismissed as ordinary inconvenience, which employees feel entitled to raise them, and whether unresolved patterns are visible to decision makers. [7] [16]

A six-step discipline for an existing leadership huddle

The following framework expands the practical checklist in Carvalho’s original manuscript. It is intended for one defined workflow concern within an existing operational meeting. It should complement clinical governance, incident reporting, and formal escalation arrangements. AHRQ’s TeamSTEPPS debrief guidance offers a related approach to reviewing team performance and incorporating learning into future plans; it does not validate the six steps below as a package. [G1]

1. Define the exposure

Write a short problem statement that identifies the process, the affected group, and the consequence that needs attention. Replace “handover is poor” with a description of where information becomes unavailable and who needs it next. Separate the safety objective from the current method used to achieve it. A method may be redesigned while the essential safety requirement remains intact.

Record what is known, what is reported but unverified, and what must be checked. Decide whether the concern requires immediate escalation before discussing a pilot. The accountable clinical or operational leader should determine the relevant safety boundaries. A routine management meeting should never become a waiting room for a serious hazard. Avoid assigning a risk rating simply because a senior person is impatient for a decision.

2. Hear the people doing the work

Identify the people who perform the process, receive its output, and absorb the consequences when it fails. Include relevant shift patterns and professional groups. Invite a description of an ordinary workday, not only an exceptional incident. Ask where the proposed change adds effort, where information is duplicated, and where work is delayed. These questions make the operational assumptions available for examination.

Offer routes for contributions outside the meeting. A short written note, an accessible conversation, or an existing confidential reporting channel may be more appropriate than speaking before a large group. Explain how private information will be handled and the limits of confidentiality. Do not promise anonymity in a small team when details could identify someone. Record the issue in a way that supports action without unnecessary personal information.

3. Assign decision rights

Name an operational owner and, where the concern crosses boundaries, an executive sponsor. The owner coordinates the investigation and maintains the record. The sponsor resolves authority or resource barriers that the owner cannot settle. Specify who can approve the test, amend the workflow, or pause it. Responsibility should attach to an identifiable role and person rather than to a committee name alone.

Make dependencies visible. If an information technology team must change a field, a clinical leader must approve a protocol, or a workforce leader must review scheduling implications, record those dependencies and their decision points. An owner without access to the required authority cannot reliably deliver closure. Where disagreement remains, document the escalation route and the interim arrangement instead of leaving staff to infer what they should do.

4. Agree a bounded test

Choose a manageable scope and a stated purpose. Describe what will change, where it will change, who will participate, and how the team will distinguish the new process from the current one. Collect a baseline where feasible, using the same operational definitions planned for follow-up. Avoid selecting only the easiest shift if the intended process must also function during evenings, weekends, or peak demand.

Select a small set of process, outcome, and balancing measures relevant to the local question. For a handoff concern, these might include completeness of required information, delays in receiving it, and time spent completing the process. Clinical leaders should choose any patient-safety measures and stopping criteria. A practical test can inform a local decision without being described as proof that the intervention works in every setting.

5. Set two clocks

The first clock is the immediate safety check. Specify who receives urgent concerns, what requires a pause, and which established escalation route applies. The second clock is the planned review date, when the team considers accumulated evidence and decides whether to continue, adapt, or stop. These clocks serve different purposes. A scheduled review must never defer a response that is needed now.

Set deadlines that reflect risk and operational reality. If information is missing at the review date, document what remains unknown, who will obtain it, and when the next decision is due. Repeated extensions should be visible to the sponsor. A date is useful because it creates an expectation of a response, not because it guarantees that a complicated problem will be resolved on schedule.

6. Close the loop

Return a clear account of the concern, the decision, the reason, the owner, and the next step. Explain which suggestion was adopted and which was not. Distinguish a completed administrative action from a resolved operational problem. Updating a policy document does not demonstrate that staff received it, understood it, or could use it under working conditions.

Confirm that the response reached the relevant people, including those who were absent when the decision was announced. Provide a route to challenge an incorrect assumption or report unintended effects. When confidentiality prevents a detailed response, explain the process and any practical change that can be shared. Closing the loop means making the disposition understandable; it does not require disclosing personal or investigative details.

Handoffs require more than a conversation

Research on structured handoff programs illustrates why implementation detail matters. Starmer and colleagues’ multicenter 2014 study evaluated an I-PASS bundle, not a slogan or isolated meeting. The program combined structured communication with training and implementation support. The reported improvement in medical errors should be interpreted in that specific intervention and study context; it is not evidence that Carvalho’s executive checklist produces the same outcome. [15]

A later prospective implementation study involved 32 hospitals, including community and academic settings. It paired an enhanced I-PASS program with longitudinal coaching and metric review. Reported handoff-related adverse events and observed communication measures improved after implementation. The design, reporting measures, implementation support, and authors’ disclosed commercial interests are relevant when assessing transferability. An executive should examine the full program and local implementation needs before borrowing a headline effect size. [5]

For this framework, the lesson is narrower: define the work and support its adoption. A revised checklist can fail if it is unavailable at the point of use, duplicates another record, or assigns tasks to someone who lacks time or authority. Ask the local team to demonstrate the process before considering wider rollout. A demonstration often exposes assumptions that remain hidden in a written policy.

Huddles need a purpose, an owner, and an evaluation plan

Franklin and colleagues’ systematic review and Pimentel and colleagues’ scoping review describe a varied huddle literature. The latter included 158 studies and identified differing structures, purposes, and measures. These reviews support treating implementation context as part of the intervention. They do not establish that every short meeting improves patient safety or that meeting attendance is an adequate outcome measure. [12] [9]

A 2024 perioperative improvement report by Tuyishime and colleagues describes a daily safety huddle in a specific service context. Such reports can offer operational ideas, while their setting and design constrain generalization. Executives should ask whether a proposed huddle has a distinct function, whether the right decision makers attend, and whether the process produces a usable disposition for the issues raised. [2]

Consider retiring or combining meetings that duplicate the same information without changing a decision. Protect a concise agenda: new exposure, unresolved actions, decisions needed, and responses due. Detailed problem solving can occur in a smaller working group with appropriate expertise. The huddle should connect the work to action rather than require every attendee to remain through a lengthy discussion that does not concern them.

Measure follow-through without confusing activity with impact

Build a small measurement set around the actual decision. A process measure might track the proportion of material concerns assigned an owner. A response measure might track whether an explanation was returned by the agreed date. An outcome measure should reflect the workflow being improved. A balancing measure should look for added workload, delays, or another unintended consequence. These categories organize local learning; they are not a validated score.

Specify denominators and exceptions before reporting results. “Ninety percent closed” is ambiguous if duplicate concerns were removed differently each week or closure means only that an email was sent. Define which concerns enter the log, how related reports are linked, and what evidence supports closure. Retain unresolved issues and explain exclusions. Review a sample of closed items with staff to learn whether the documented resolution matches their experience.

Use established culture tools when appropriate. AHRQ’s Hospital Survey 2.0 includes domains covering communication openness, response to error, management support, and handoffs. Follow the instrument’s administration and interpretation guidance. Locally invented pulse questions may be useful for discussion, but should not be presented as equivalent to a validated instrument or as proof that a management change caused better clinical outcomes. [G2]

Interpret increased reporting carefully. Combine counts with the type of concern, exposure to the underlying workflow, independent safety information, and staff accounts. A rising count could reveal previously hidden problems. A falling count could reflect safer work or reduced willingness to speak. Before celebrating either direction, ask what changed in both the work and the reporting process. The 2025 evidence review makes this ambiguity particularly important. [1]

Feedback is part of the leadership work

Sexton and colleagues’ cross-sectional study of Leadership WalkRounds found that feedback about actions was associated with more favorable safety-culture and engagement measures. It did not establish whether feedback caused those differences or reflected stronger organizational conditions. Tawfik and colleagues likewise reported associations between local leadership behavior and workforce well-being, safety culture, and engagement. These findings support examining leadership conduct while avoiding causal promises about a single tactic. [14] [4]

Make feedback routine enough that it does not depend on personal memory. Use the same short structure for adopted, declined, and deferred suggestions. Explain resource constraints honestly. When a team cannot act because it lacks authority, send that barrier upward with an explicit decision request. An employee should not have to submit the same concern repeatedly simply to discover where responsibility sits.

Feedback should also travel toward leadership. Invite staff to say whether the explanation was understandable and whether the revised process works. If the answer contradicts the official status, reopen the issue. A management process earns credibility when it can correct its own account. The aim is a reliable connection between decision makers and working conditions, rather than a record that makes every action appear complete.

A hypothetical handover example

Consider a hypothetical adult social care service in England introducing an additional handover step. Staff report that its timing conflicts with peak care activity. This is a composite illustration, not a documented client case. No measured improvement, financial return, or patient outcome is claimed. Its purpose is to show how a manager could retain a safety objective while questioning the sequence used to achieve it.

The manager first establishes the information that must reliably reach the next responsible colleague. Staff describe when the proposed step becomes difficult and which roles depend on its output. The manager assigns an operational owner, asks the relevant professional lead to confirm safety boundaries, and selects one limited test. The team records incomplete information, delays, time required, and any concerns that need immediate escalation.

At the planned review, the manager considers the record alongside staff accounts. A positive staff reaction alone would not establish safety. Equally, a completed form would not establish that important information reached its recipient. The decision might be to revise sequencing, retain part of the original method, extend observation, or stop the test. Whatever the decision, staff receive the reason and the next review point.

For US hospital leaders, the transferable idea is the decision discipline. The staffing model, clinical accountabilities, technology, and regulatory context require local assessment. A US organization should not import an English adult social care workflow or treat this illustration as evidence of effectiveness. Local clinical, operational, workforce, privacy, and compliance leaders should determine how any proposed process fits their setting.

A practical first 30 days

In the first week, choose one recurring workflow concern with a manageable scope. Review existing reporting and escalation routes before adding another channel. Agree the problem statement, owner, sponsor, safety boundaries, and staff groups that should contribute. Establish what the team can reasonably learn in a month and what will require longer observation. Document these choices in plain language.

During the second week, map the current process with the people who use it and prepare the bounded test. Confirm how data will be collected and how confidentiality will be protected. Show participants how to raise an urgent issue and how to record routine feedback. Ask someone unfamiliar with the plan to explain it back; confusion at this stage is useful information about the design.

In the third week, review early experience and implementation fidelity. Determine whether the process was actually used, by whom, and under what conditions. Inspect burdens and workarounds as well as intended results. Correct urgent problems through established channels. Avoid widening the test simply because the initial response is enthusiastic. The purpose is to learn enough to make a responsible decision.

In the fourth week, make and communicate the next decision. State the evidence considered, the limitations, the owner of any remaining work, and the next review date. Decide which elements should stop, continue, or change. Keep a short record that another manager could understand without attending every meeting. Treat a well-supported decision to stop an impractical change as a useful result, not a failure to demonstrate progress.

Make the values visible

The Healthcare Executive’s values include integrity, innovation, collaboration, excellence, inclusivity, and service. This framework translates those commitments into observable management behavior. Integrity means explaining the evidence and its limits. Innovation means testing an idea with appropriate safeguards. Collaboration means involving the people whose work is affected. Excellence means checking implementation and outcomes. Inclusivity means considering whose experience is absent. Service means keeping the effects on people and communities in view.

A useful leadership review can end with five questions: What did we hear? Who owns the decision? What happens if the risk cannot wait? What evidence will we review? How will people learn the outcome? The questions are simple, but answering them consistently requires attention to authority, working conditions, and the quality of the response. That is where a humane leadership intention becomes a practical management commitment.

Author and editorial context

Mafalda Carvalho is a Portuguese-born author, speaker, and founder of Prestige Care Consulting, based in Greater Manchester, UK. Her experience includes management of adult social care services in England, quality assurance, safeguarding, workforce development, and the practical implementation of change. She holds a BSc (Hons) in Health and Care Management from Arden University. She is the author of The Prestige Leader: The Human Side of Healthcare Leadership.

Commercial interest: Carvalho is the founder of Prestige Care Consulting. This contributed perspective is not a product endorsement or a report of independently evaluated consultancy outcomes. The Healthcare Executive expanded the literature discussion and practical application material for Management Atlas. References were checked against publisher, author-repository, or PubMed records on October 25, 2026. This is a focused narrative evidence review, not a systematic review or a comprehensive literature search.