Diagnostic pause: stop before closing the case
Table of contents
It is four in the afternoon and the patient in bed seven has been in the hospital for two days with a diagnosis that everyone repeats on rounds. The prescription has been adjusted, discharge is planned for early tomorrow, and the progress note has already been copied twice. During handover, the resident asks a question no one had asked out loud: "if it is not this, what else could it be?". The conversation lasts three minutes. Someone remembers that the fever did not subside as expected, another notices that an admission test was never reviewed. In the end, the team requests an evaluation that changes the story.
This question, asked on purpose and at a set moment, has a name: diagnostic pause, or diagnostic time-out. It does not depend on technology, does not require a meeting and fits within an already tight routine. At the same time, it is not magic, and studies show clear limits. This guide explains what it is, where the idea came from, what research has found so far, when it is worth stopping, how to put together a short script and how to bring the habit into the internship and residency without turning everything into bureaucracy.
Key points
- The diagnostic pause is a set moment to review the main hypothesis before closing the case, explicitly asking what else could explain the picture.
- It comes from the idea of a cognitive forcing strategy: a simple rule that forces reasoning out of autopilot at points of risk.
- Premature closure, when the team stops considering alternatives after reaching a first hypothesis, is one of the most cited cognitive causes of diagnostic error.
- Guided reflection has the best evidence among cognitive interventions; isolated checklists and "debiasing" training have mixed results.
- The pause works best at defined triggers (transfer, discharge, unexpected worsening, a sense that something does not add up) than as a ritual for every patient.
- In teaching, the pause is an opportunity to show reasoning out loud and to give the student permission to disagree.
Contents
- What a diagnostic pause is
- Why reasoning closes too early
- Where the idea comes from
- What studies show
- When to take the pause
- A five-question pause script
- How to bring the pause into teaching
- How NEXO helps train hypothesis review
- Frequently asked questions
- Diagnostic pause checklist
What a diagnostic pause is
A diagnostic pause is a short, planned and explicit interval in which you, alone or with the team, stop executing the plan and look again at the diagnosis guiding care. The central question is simple: does the hypothesis we are following explain all the data, or are we ignoring something that does not fit? The difference between this pause and ordinary clinical thinking lies in the adjective "explicit". Every doctor reflects on their patients, but does so irregularly, when there is time left or when something draws attention. The pause turns this reflection into a commitment: it happens at a defined point, follows a script and ends with a recorded decision.
The term comes from a parallel with the surgical pause, that moment before the incision when the team confirms patient, side and procedure. In the operating room, the pause protects against identification errors. On the ward, in the emergency department or in the ICU, the diagnostic pause protects against another kind of failure: moving forward with a hypothesis that was never tested against alternatives. Yale, Cohen and Bordini describe the diagnostic time-out precisely as a cognitive forcing function, a mechanism that forces the clinician to switch from fast mode to analytical mode at a chosen moment.
What the pause is not
The diagnostic pause is not a formal second opinion, nor a one-hour case discussion, nor an audit. Nor is it general distrust of the team that made the initial diagnosis. Those who reach a quick hypothesis are, most of the time, right, and trained intuition is one of the most valuable tools in clinical practice. The pause only asks that this hypothesis go through a short test before becoming fixed truth in the chart.
The three elements of a good pause
A useful pause has a trigger (the moment it happens), a script (the questions that will be asked) and an outcome (what changes, or the conscious decision not to change). Without a trigger, the pause becomes a vague intention. Without a script, it becomes loose conversation that depends on who is on shift. Without an outcome, it becomes a ritual, and rituals without consequence are the first to disappear when the service gets busy.
Why reasoning closes too early
To understand why it is worth stopping, you need to understand why reasoning tends to close early. Medicine works with a dual-process model: a fast, intuitive system based on pattern recognition, and a slow, analytical system that compares hypotheses and weighs evidence. Both are necessary. The fast one lets you handle a crowded emergency department; the slow one comes in when something does not fit. The problem is that the fast system does not warn you when it is wrong. It delivers an answer with a feeling of certainty, and that feeling is the same whether the answer is right or wrong.
Premature closure
In the classic study by Graber, Franklin and Gordon, which analyzed one hundred cases of diagnostic error in internal medicine, cognitive factors contributed to the error in 74% of cases and system factors in 65%, and premature closure, defined as failing to consider reasonable alternatives after reaching an initial diagnosis, was the most common single cause. This finding matters for two reasons. First, it shows that the problem is rarely a total lack of knowledge: the doctor knew the right disease, but stopped looking before reaching it. Second, it shows that diagnostic error almost never has a single cause. System and cognition mix, and a pause that ignores the context (tests that have not come back, rushed shift handovers) loses half its value.
Availability bias and diagnostic momentum
Two other mechanisms feed early closure. Availability bias makes the hypothesis that comes to mind most easily seem more likely, usually because you saw a similar case recently. Diagnostic momentum is the inertia of a label: once written at admission, it passes from one professional to another, gains weight with each repetition and becomes harder and harder to question. In an experiment with internal medicine residents, Mamede and colleagues showed that recent experience with similar cases induced availability bias in second-year residents and that structured reanalysis of the findings corrected much of that error. The diagnostic pause is, in essence, a way of bringing that structured reanalysis to the bedside.
Bias or lack of knowledge?
There is an honest debate in the literature about the relative weight of biases and lack of knowledge. The review by Norman and colleagues argues that errors arise from both the intuitive and the analytical system and that, with more experience and more knowledge, the probability of error falls. For those who teach, this has a practical consequence: a diagnostic pause made by someone who does not know the alternatives does not generate alternatives. That is why the script works better when it comes with study of the differential diagnosis, such as the one described in the guide on how to train differential diagnosis with clues.
Where the idea comes from
The idea of stopping at defined points to review reasoning has roots in two traditions. The first is patient safety, which borrowed from aviation the logic of the checklist and cross-checking. The second is cognitive psychology applied to medicine, which began to describe errors of reasoning with the same care used to describe procedural errors.
Cognitive forcing strategies
The most cited conceptual framework is Croskerry's article on cognitive forcing strategies in emergency medicine. He proposes that, through metacognition, that is, thinking about one's own thinking, the clinician can develop rules that interrupt latent errors. Croskerry describes three levels: universal strategies (knowing biases in general), generic ones (applicable to common situations, such as "always consider the worst-case scenario") and specific ones (tied to concrete pitfalls of particular conditions). The diagnostic pause is a generic strategy: it applies to any patient, as long as the trigger is activated.
From pediatrics to the ICU transition
In recent years, the idea has gained practical versions. In hospital pediatrics, Yale and colleagues adapted a diagnostic time-out and tested it for twelve months with a group of eight physicians; in more than half of the cases, the pause did not confirm the initial diagnosis and other hypotheses were investigated, and in every case some new action was taken. In a pediatric emergency department, Kasick and colleagues used a pause to improve documentation of the differential diagnosis in children admitted with abdominal pain. Cornell and colleagues studied a handoff tool that builds a diagnostic pause into the transfer from the ICU to the ward, and the intensivists interviewed recognized the space to discuss uncertainty as the main benefit.
.jpg){kind=link}
What the studies show
The question every service asks before adopting a practice is whether it works. For the diagnostic pause, the honest answer is: the evidence is promising on some points, weak on others, and almost all of it comes from small studies. It is worth a careful look.
Guided reflection has the best evidence
The systematic review with meta-analysis by Prakash, Sladek and Schuwirth gathered 44 studies on cognitive interventions to improve diagnostic decision-making. Among all the strategies, guided reflection was the only one with enough data for meta-analysis and showed a moderate effect on accuracy. Immediate feedback and modeling reflection with contrasting examples also seemed to help, but methodological problems prevented a quantitative estimate. In Brazilian students, Fernandes and colleagues showed that adding structured guidance to deliberate reflection improved diagnostic accuracy in cases of different levels of complexity, in both the third and the sixth year.
Checklists have mixed results
Diagnostic checklists are the most concrete form of pause, and results vary. In the pilot trial by Ely and Graber, with fourteen primary care physicians, the error rate with a checklist was not significantly different from that of usual care (11.2% versus 17.8%), although the physicians with the checklist considered more diagnoses. With Japanese students, Shimizu, Matsumoto and Tokuda found improvement with a differential diagnosis checklist in difficult cases, but worsening in simple cases, and no gain with a generic bias checklist. In electrocardiogram interpretation, Sibbald and colleagues found no overall advantage for either bias checklists or knowledge checklists, with a possible benefit for novices.
Teaching biases alone is not enough
A short review by Oliver, Oliver and Body concluded that there is little evidence that teaching cognitive forcing strategies, by itself, reduces cognitive errors in emergency medicine. This does not invalidate the pause, but it shifts the focus: instead of memorizing lists of biases, it is worth training the concrete action of reviewing hypotheses in everyday cases and in study cases.
| Intervention | What was tested | Main result | What to take into practice |
|---|---|---|---|
| Guided reflection | Structured reanalysis of findings in written cases | Moderate improvement in accuracy in meta-analysis | Basis of the pause script |
| Reflection with guidance | Step-by-step instructions for students | Better accuracy on the immediate test | Give ready-made questions to the novice |
| Differential checklist | List of diagnoses by complaint | More hypotheses considered; effect on error uncertain | Use in difficult cases, not in all |
| Bias checklist | Generic list of cognitive biases | No consistent gain | Prefer concrete questions about the case |
| Pause in pediatrics | Script applied for twelve months | Change in management in all cases analyzed | Choosing cases by clinical intuition |
| Pause at the ICU transition | Shift handover tool | Perceived benefit; workflow and culture barriers | Embed the pause in the process that already exists |
When to do the pause
If the pause happens for every patient, every day, it will become a rubber stamp in less than a week. The pediatrics study by Yale and colleagues showed that the most chosen cases were those in which the clinician's intuition said something did not add up. The triggers below work well as a starting point for a service or for your own habit.
- Care transitions: admission, transfer between units and shift handover, when the diagnostic label tends to be copied without review.
- Before discharge: the last moment when the team still has the patient nearby to correct an error.
- Worsening or unexpected course: when the patient does not respond to treatment as expected for the assumed diagnosis.
- Early return: the patient who comes back to the emergency department within a few days with the same complaint.
- Data that do not fit: a test outside the pattern, a sign that no one explained, a history that changed.
- Discomfort of the professional: the feeling, yours or someone on the team's, that something is wrong, even without knowing what.
How long it takes
The main barrier cited by professionals is time. In the survey with neonatal ICU professionals described by Yale, Cohen and Bordini, the perceived lack of time appeared as a universal barrier. In practice, a pause with a ready script usually fits in a few minutes. In the pediatrics study, only one case was considered time-consuming. The secret is to have the questions ready and to know in advance who leads.
Alone or as a team
The pause can be individual, before signing a discharge or writing a progress note, or collective, at rounds or at shift handover. The team version has an advantage: different people notice different data, and nursing, which spends more time with the patient, often brings the missing information. The guide on clinical reasoning in the internship shows how to organize hypotheses by priority, which makes the conversation easier.
A five-question pause script
The script below brings together elements that appear in published versions of the diagnostic pause and in deliberate reflection studies. It was designed to fit in a few minutes and to be led by anyone on the team, including the student.
- What is the main hypothesis and why do we believe in it? Say aloud the data that support the diagnosis. If the justification is "that is what came in at admission", that is already a sign.
- What data do not fit? Actively look for the finding that the hypothesis does not explain: a test, a vital sign, a complaint that was set aside.
- If it is not this, what else could it be? List at least two alternatives, including the most serious one that cannot go unnoticed.
- What data would separate the hypotheses? Think of the test, the history question or the observation time that would change the probability of each one. The guide on pretest probability helps at this step.
- What do we do now? Decide and record: maintain, expand the investigation, ask for help, or agree on when to reassess.
How to record it in the chart
One line is enough: "Diagnostic pause at transfer: hypothesis maintained; alternatives considered X and Y; reassess if Z". This record has two effects. It tells the next shift that the hypothesis was tested, and it shows that the alternatives were considered, which helps whoever arrives not to repeat the same conversation from scratch. The guide on how to organize clinical reasoning in the chart offers templates for writing the assessment.
Common mistakes when applying the script
The first mistake is turning the pause into an interrogation of whoever made the diagnosis. The question is about the case, never about anyone's competence. The second is stopping at the list of alternatives without deciding anything; a pause with no outcome teaches the team that the conversation is useless. The third is applying the same pause to every patient, instead of using triggers. The fourth is leaving the pause only to the supervising physician, when the student and the nursing staff are precisely the ones who most often bring new data.
How to bring the pause into teaching
The diagnostic pause is a safety tool and also a teaching tool. It makes visible a process that usually happens in the experienced physician's head and that the student rarely sees. When the supervising physician stops and says "let's do the pause", they are showing, in real time, how a hypothesis is tested.
Reasoning out loud
Ask the intern to lead the pause and say out loud the data that support the hypothesis, the ones that do not fit and the alternatives. Then add what you saw. This format draws on what the deliberate reflection literature found: the student gains when given structured questions, and not just the generic instruction to "think better". Reflection with guidance was more useful than free reflection in studies with students.
Permission to disagree
The biggest obstacle to the pause in a team is hierarchy. The intern who notices a strange datum may not speak for fear of seeming unsure. Agree from the first day that the question "what else could it be?" is welcome from anyone, and thank the person when it comes up, even if the hypothesis is kept. The text on metacognition in medicine discusses how to teach the student to monitor their own reasoning, and the guide on diagnostic uncertainty helps with talking about what is not yet known.
Training outside the shift
The pause becomes easier on the shift when it has already been trained in study cases. Written cases, discussion sessions and diagnosis games allow the act of reviewing the hypothesis to be repeated many times, with no risk to anyone. What matters is that each case ends with the same question: what made me keep or change the hypothesis?
How NEXO helps train hypothesis review
In NEXO, the case of the day arrives in up to six clues, and each new clue works as a small pause: you can keep the hypothesis or change it in light of the datum that has just appeared. With each guess, the proximity in ICD-10 across five levels shows how far your hypothesis was from the diagnosis, which helps you notice whether you were in the right chapter or whether the reasoning closed too early on a wrong path.
The cases are fictional with a real clinical presentation, and the guide what NEXO is explains the whole game. How the modes are unlocked is on the NEXO economy page. A good exercise is to write down, before each clue, the two alternatives you would keep if the main hypothesis were wrong, and then compare with the outcome.
If you coordinate a medical school, a student league or a residency and want to train the pause in a group, see NEXO Closer. Researchers in medical education can propose studies through NEXO Researcher. Physicians who review cases take part through NEXO Curator, and those who create health content on social media can take part through NEXO Creators.
Frequently asked questions
Is the diagnostic pause the same thing as the surgical pause?
No. The surgical pause confirms patient, side and procedure before the incision. The diagnostic pause reviews the hypothesis that is guiding care. Both share the idea of stopping at an agreed point, but they protect against different errors.
Do I need to do the pause with every patient?
It is not recommended. The pause works best with defined triggers, such as transfer, discharge, unexpected worsening or data that do not fit. Applied to everyone, it tends to become a formality.
Does the diagnostic pause reduce errors?
Guided reflection, which is the basis of the pause, improved diagnostic accuracy in a meta-analysis. The specific studies on the pause are small and show a change in conduct and a perceived benefit, but there is still no solid proof of error reduction in patient outcomes.
How long does a pause take?
With a script ready, a few minutes. Lack of time is the most cited barrier, but in the pediatrics study only one case was considered time-consuming.
Does a checklist replace the pause?
Differential diagnosis checklists can help in difficult cases, but the results are mixed and they can get in the way in simple cases. They work better as support for a script of questions than as a substitute.
Can the student call for the pause?
They can and should. Agreeing that anyone on the team can ask "what else could it be?" is one of the simplest ways to make the pause part of the service's culture.
Diagnostic pause checklist
- Define your service's triggers: transfer, discharge, unexpected worsening, early return, data that do not fit.
- Agree on who leads the pause and make it clear that anyone can ask for one.
- Use the five questions: hypothesis and justification, data that do not fit, alternatives, the finding that separates them, decision.
- Always include the most serious alternative that cannot go unnoticed.
- Record one line in the chart with the hypothesis, the alternatives and the criterion for reassessment.
- Practice reviewing hypotheses in study cases before requiring the pause on shift.
- From time to time, review whether the pause is changing conduct or has become a rubber stamp.
Want to practice reviewing hypotheses with every clue? Play the case of the day on NEXO.
References
- Yale S, Cohen S, Bordini BJ. Diagnostic Time-Outs to Improve Diagnosis. Crit Care Clin. 2022. PubMed 35369942
- Croskerry P. Cognitive forcing strategies in clinical decisionmaking. Ann Emerg Med. 2003. PubMed 12514691
- Yale SC, Cohen SS, Kliegman RM, Bordini BJ. A pause in pediatrics: implementation of a pediatric diagnostic time-out. Diagnosis (Berl). 2022. PubMed 35417931
- Kasick RT, Melvin JE, Perera ST, Perry MF, Black JD, Bode RS, et al. A diagnostic time-out to improve differential diagnosis in pediatric abdominal pain. Diagnosis (Berl). 2021. PubMed 31677376
- Ely JW, Graber MA. Checklists to prevent diagnostic errors: a pilot randomized controlled trial. Diagnosis (Berl). 2015. PubMed 29540029
- Oliver G, Oliver G, Body R. BET 2: Poor evidence on whether teaching cognitive debiasing, or cognitive forcing strategies, lead to a reduction in errors attributable to cognition in emergency medicine students or doctors. Emerg Med J. 2017. PubMed 28724568
- Fernandes RAF, Malloy-Diniz LF, de Vasconcellos MC, Camargos PAM, Ibiapina C. Adding guidance to deliberate reflection improves medical student's diagnostic accuracy. Med Educ. 2021. PubMed 33978967
- Cornell EG, Harris E, McCune E, Fukui E, Lyons PG, Rojas JC, et al. Scaling up a diagnostic pause at the ICU-to-ward transition: an exploration of barriers and facilitators to implementation of the ICU-PAUSE handoff tool. Diagnosis (Berl). 2023. PubMed 37598362
- Mamede S, van Gog T, van den Berge K, Rikers RM, van Saase JL, van Guldener C, et al. Effect of availability bias and reflective reasoning on diagnostic accuracy among internal medicine residents. JAMA. 2010. PubMed 20841533
- Shimizu T, Matsumoto K, Tokuda Y. Effects of the use of differential diagnosis checklist and general de-biasing checklist on diagnostic performance in comparison to intuitive diagnosis. Med Teach. 2013. PubMed 23228085
- Sibbald M, Sherbino J, Ilgen JS, Zwaan L, Blissett S, Monteiro S, et al. Debiasing versus knowledge retrieval checklists to reduce diagnostic error in ECG interpretation. Adv Health Sci Educ Theory Pract. 2019. PubMed 30694452
- Graber ML, Franklin N, Gordon R. Diagnostic error in internal medicine. Arch Intern Med. 2005. PubMed 16009864
- Prakash S, Sladek RM, Schuwirth L. Interventions to improve diagnostic decision making: A systematic review and meta-analysis on reflective strategies. Med Teach. 2019. PubMed 30244625
- Norman GR, Monteiro SD, Sherbino J, Ilgen JS, Schmidt HG, Mamede S. The Causes of Errors in Clinical Reasoning: Cognitive Biases, Knowledge Deficits, and Dual Process Thinking. Acad Med. 2017. PubMed 27782919