CRP03 — Differential diagnosis under uncertainty

ClinicalEdu V1.14 | Adult NHS educational reading | Editorial review

Learning context and original foundational reading

Differentials are tools for action

A differential diagnosis is not a prize for producing the longest list. Its purpose is to manage risk while uncertainty remains. Organise candidate explanations by urgency, plausibility and treatability. One category covers immediately threatening mechanisms to assess now, another covers likely explanations, and a third covers alternatives that become important if the expected response fails or new evidence emerges.

Do not confuse likelihood with consequence

A relatively uncommon but catastrophic possibility can deserve immediate evaluation even when a benign diagnosis is more probable. Likewise, a common cause such as dehydration can coexist with infection, bleeding or a medication adverse effect. Explicitly name the “must not miss” mechanism and the observations or investigations that would help separate it from less urgent alternatives.

Cognitive bias as a practical risk

Anchoring appears when a plausible early explanation prevents further interpretation. Confirmation bias appears when only supportive results are remembered. Premature closure appears when a label ends enquiry despite mismatching physiology. These concepts are useful when translated into actions: ask what does not fit; seek a second explanation; repeat observations; invite a colleague to challenge the model; and record what result would prompt a change in plan.

Test alternatives actively

A differential should generate discriminating questions rather than generic testing. For each diagnosis ask: what should I find if this is true? Which clinical finding would make it less credible? What is the safest next test? Would a positive result actually change management? A diagnostic test can cause delay, transfer or treatment harm, so its value must be judged against the decision at hand.

Set review points

Uncertainty is safe only when accompanied by monitoring, escalation criteria and ownership. Record the working diagnosis as provisional, specify the alternative concern and define a time-bound reassessment. If the patient is deteriorating, repeat examination and seek senior support rather than merely ordering additional tests. A change in physiology can make the original differential obsolete.

Worked clinical scenario

An older inpatient becomes hypoxic, restless and tachycardic after a recent procedure. Construct an urgent-first differential, including mechanisms that coexist, and define which finding triggers immediate help.

Constructing and testing a useful differential

The purpose of a differential diagnosis is to guide immediate decisions, not to demonstrate encyclopaedic recall. Construct a small set of plausible causes and a parallel set of conditions that are less likely but dangerous to miss. An explanation can be common without being sufficient. In an older person who is breathless, known heart failure may be salient, but pulmonary embolism, infection, acute coronary pathology or medication-related respiratory depression may need consideration depending on features and context.

Use an explicit “what would change my mind?” question. For each working hypothesis, identify findings that support it, findings that weaken it and a next discriminating observation or investigation. Beware diagnostic momentum: once a label appears in a referral or handover, later clinicians may unconsciously treat it as confirmed. Avoid treating an abnormal test as independent confirmation if it was ordered because of the very suspicion it now appears to validate.

Cognitive bias is most useful when made operational. Anchoring can be countered by a deliberate alternative; premature closure by noting outstanding risks; availability bias by examining base rates and salient but uncommon diagnoses; confirmation bias by looking for disconfirming evidence. The goal is not to eliminate all heuristics—they are necessary in time-limited work—but to insert checkpoints where error would be costly. Document the plan for reassessment when uncertainty remains. A safe unresolved differential is often preferable to a confident incorrect diagnosis.

Extended worked cases and reasoning prompts

Case discussion 1

A patient receiving anticoagulation falls, then appears increasingly confused. An early note calls this “mechanical fall with delirium”. What dangerous alternative would you actively consider, and which examination findings and urgent actions are needed? Distinguish an explanatory working diagnosis from a decision to seek emergency imaging or senior advice.

Case discussion 2

Now consider a patient with fever, tachycardia and a painful red leg. Outline likely and high-consequence alternatives rather than listing every infectious disease. State which questions discriminate and when treatment or transfer cannot wait.

Structured practice and reflection

1. Write a two-sentence problem representation separating observed facts from assumptions.

2. Name two important alternatives and one potentially harmful missed condition.

3. What is the safest immediate action before a definitive diagnosis is available?

4. Which finding or change would cause you to revise your plan?

5. What would you document or communicate to make responsibility and follow-up explicit?

Further reading — primary organisations

NICE acutely ill patients CG50: https://www.nice.org.uk/guidance/cg50

GMC Good medical practice: https://www.gmc-uk.org/professional-standards/the-professional-standards/good-medical-practice

Editorial verification: only selected central documents have been checked live this session; the complete module reference list requires individual date/version checking and clinical reviewer sign-off before public release.

Educational boundary: an online completion record is not evidence of independently assessed clinical competence or accredited BLS/ALS provider certification.

Clinical decision and professional scope

Calibration rather than premature closure is the central theme of this chapter. The case begins with: patient with breathlessness, fever absent, newly tachycardic and several possible explanations. The educational aim is to practise structured reasoning, seek timely senior help and revisit early judgements as the patient changes. It is not a substitute for individual assessment, escalation policy, approved practical training or independent clinical supervision.

Reasoning in the first clinical encounter

Build a prioritised differential, not a long unranked catalogue. Ask which alternative would be dangerous to miss, which is most likely, and what finding would distinguish them.

Ask what must be recognised now, what is uncertain, which risks cannot wait and what professional or practical resources are available. This is especially important in the following scenario: Patient with breathlessness, fever absent, newly tachycardic and several possible explanations.

Evidence, bias and contradictory information

Avoid anchoring on the first ward diagnosis. New hypoxia may be due to more than infection; initial uncertainty is a reason for planned reassessment, not a reason to delay escalation.

Distinguish reliable measurements, recalled history, inferred mechanisms and assumptions. Identify a genuine contradictory fact and explain how it changes the plan. This is especially important in the following scenario: Patient with breathlessness, fever absent, newly tachycardic and several possible explanations.

A worked bedside narrative

Explicitly document what evidence would make you change your mind and when that evidence should be available.

Treat the first impression as provisional. Explain the immediate actions, an alternative explanation and the findings that would determine escalation. This is especially important in the following scenario: Patient with breathlessness, fever absent, newly tachycardic and several possible explanations.

Follow-up and safety net

Build a prioritised differential, not a long unranked catalogue. Ask which alternative would be dangerous to miss, which is most likely, and what finding would distinguish them.

Specify who repeats observations, the interval dictated by local protocols and what will trigger additional clinical review. This is especially important in the following scenario: Patient with breathlessness, fever absent, newly tachycardic and several possible explanations.

Communication and record quality

Avoid anchoring on the first ward diagnosis. New hypoxia may be due to more than infection; initial uncertainty is a reason for planned reassessment, not a reason to delay escalation.

Show that the documentation allows a colleague to see the decision, uncertainty, response and accountable next action. This is especially important in the following scenario: Patient with breathlessness, fever absent, newly tachycardic and several possible explanations.

Teaching exercise

Explicitly document what evidence would make you change your mind and when that evidence should be available.

Pause at each stage to write a short problem representation and a concrete decision before looking at the proposed reasoning. This is especially important in the following scenario: Patient with breathlessness, fever absent, newly tachycardic and several possible explanations.

Extended case discussion

Initial contact: Patient with breathlessness, fever absent, newly tachycardic and several possible explanations. The clinician has incomplete handover details. Record the physiological observations, compare the current presentation with baseline and decide whether urgent response is warranted. This stage is about recognising the need for action, not guessing the final diagnosis.

At the first review: Ask whether the patient can supply a reliable history; consult relevant carers and records; reconcile medication changes; and interpret findings in the context of frailty and multimorbidity. The team must not allow uncertainty about cause to hide objective deterioration.

New information: A colleague reports that the change began earlier than originally thought. The clinician must revise the time course and identify whether this shifts the prioritised differential, appropriate investigation, monitoring frequency or destination of care.

Decision point: Rank time-critical causes without claiming false certainty. Explain a decision that is justified now, a competing option and the circumstances in which it would be appropriate. Clearly distinguish what you know, what you suspect and what needs confirmation.

After the intervention: Arrange reassessment and document the response. A settled score, a single laboratory result or a temporary symptom improvement should not erase the need to review the full trajectory.

Practice lens 1: time and trajectory

Build a prioritised differential, not a long unranked catalogue. Ask which alternative would be dangerous to miss, which is most likely, and what finding would distinguish them. Consider how this principle is affected by imperfect staffing, overnight cover or limited site resources; distinguish a desirable ideal response from what can actually be delivered safely.

Write a structured clinical note using the headings: observed findings; baseline; current concern; differential and important alternatives; agreed actions; accountable reviewer; and explicit reassessment or transfer triggers. Highlight any unresolved question instead of silently closing it.

Practice lens 2: uncertainty and corroboration

Avoid anchoring on the first ward diagnosis. New hypoxia may be due to more than infection; initial uncertainty is a reason for planned reassessment, not a reason to delay escalation. Consider how this principle is affected by imperfect staffing, overnight cover or limited site resources; distinguish a desirable ideal response from what can actually be delivered safely.

Write a structured clinical note using the headings: observed findings; baseline; current concern; differential and important alternatives; agreed actions; accountable reviewer; and explicit reassessment or transfer triggers. Highlight any unresolved question instead of silently closing it.

Practice lens 3: next decision and ownership

Explicitly document what evidence would make you change your mind and when that evidence should be available. Consider how this principle is affected by imperfect staffing, overnight cover or limited site resources; distinguish a desirable ideal response from what can actually be delivered safely.

Write a structured clinical note using the headings: observed findings; baseline; current concern; differential and important alternatives; agreed actions; accountable reviewer; and explicit reassessment or transfer triggers. Highlight any unresolved question instead of silently closing it.

Reading and source-check register

Core reference: NICE CG50. The reference supports its own clinical or professional recommendations; the case examples, reflective prompts and teaching structure are original educational synthesis, not direct source quotations. Check the operative full guidance and local policy before applying recommendations.

Reflective practice

What did I first assume? Which finding would most quickly falsify that assumption? Was I influenced by site capability or workload? What did I communicate, to whom, and when should that decision be reviewed? Do not include patient-identifiable information in submitted reflections.

Boundary with BLS and ALS

Early recognition, timely escalation, clinical reasoning and team communication support later BLS/ALS theoretical learning. This chapter does not teach or certify CPR, defibrillation, airway procedures or accredited resuscitation competence. Use current Resuscitation Council UK guidance and approved supervised practical training for those skills.