01Principles and purposeThe professional or clinical skill and the decisions it supports.
Severity assessment has three related jobs: recognising present instability, anticipating a difficult course and describing the illness once its evolution is known. Confusing these jobs causes delay. A score can suggest increased risk without showing established organ failure, while a patient with a modest score may already need respiratory or circulatory support. The bedside priority is to identify the current physiological problem and choose a location where it can be treated and reassessed safely.
The revised severity categories distinguish mild disease, moderately severe disease and severe disease. Transient organ failure resolves within 48 hours; failure persisting beyond that interval makes the illness severe. Local complications or an exacerbation of a comorbidity can make disease moderately severe without persistent failure. These definitions describe the observed course and may change as information accrues. An early label should therefore remain provisional, with a handover that states the actual abnormalities, their onset and the response to treatment.
Key points
- Predictive scores estimate risk; the patient’s evolving organ dysfunction determines urgent care and the eventual severity category.
- Mild disease has no organ failure or local/systemic complication; moderately severe disease has transient failure or complications without persistent failure.
- Severe acute pancreatitis is defined by organ failure persisting beyond 48 hours, but support and escalation begin when dysfunction appears.
- Trend oxygen need, circulation, mentation, urine output, fluid balance and renal biochemistry; one reassuring measurement does not establish a safe trajectory.
- SIRS is an inflammatory signal and risk marker, not proof of infection or a stand-alone antibiotic indication.
- Repeated fluid assessment must consider both inadequate perfusion and overload; rising creatinine does not automatically mean that further fluid is safe.
02Situations and prioritiesThe context, relevant information and actions that matter most.
Compare oxygen requirement, respiratory rate, work of breathing and gas exchange with the previous assessment. A saturation maintained only by progressively increasing oxygen is deterioration, not stability. Consider pancreatic inflammation, atelectasis, aspiration, pulmonary oedema and other respiratory causes; the need for escalation is determined by the physiological problem even before its precise mechanism is established.
Assess blood pressure, pulse, capillary refill, mental state and urine output together. Persistent oliguria or a rising creatinine may reflect inadequate perfusion, established kidney injury, obstruction or a competing problem. Examine fluid balance and congestion before prescribing more fluid. Document whether a previous fluid intervention actually improved perfusion and whether it produced respiratory harm.
SIRS is present with at least two of temperature below 36°C or above 38°C, heart rate above 90/min, respiratory rate above 20/min or PaCO2 below 32 mmHg, and white cells below 4 or above 12 ×10⁹/L or more than 10% immature band forms. In acute pancreatitis it supports closer monitoring and risk assessment. It is not a diagnosis of bacterial infection; sterile inflammation can produce the same pattern.
Increasing distension, pain, vomiting, gastrointestinal bleeding or new sepsis features should trigger examination for a complication or another diagnosis. A tense abdomen combined with declining urine output and difficult ventilation raises concern for intra-abdominal hypertension or compartment syndrome. Senior review should coordinate the appropriate measurement, imaging and support rather than attributing every abnormality to the original pain.
03Assessment and interpretationHow to gather information, assess the situation and recognise uncertainty.
Consider the information, its meaning and its limitations before deciding what follows.
- 01
Bedside observations and a timed fluid balance - Why
- Establish the direction of physiological change and treatment response.
- Interpretation and limitations
- Record oxygen delivery as well as saturation, actual urine volume and the interval over which it was measured. A urine output below approximately 0.5 mL/kg/hour is concerning in the adult assessment, particularly when persistent or accompanied by other abnormalities. Use repeated examination to distinguish an underfilled circulation from congestion; neither an observation score nor a balance total is sufficient alone.
- 02
Renal profile, haematocrit and electrolytes - Why
- Follow perfusion-related trends and detect treatment or disease complications.
- Interpretation and limitations
- Rising urea, creatinine or haematocrit can support concern about inadequate circulating volume, but interpretation depends on baseline kidney function, losses, intake and signs of overload. An isolated target should not drive unrestrained fluids. Potassium, calcium and glucose abnormalities need their own assessment; serial pancreatic enzymes add little to this severity monitoring.
- 03
Blood gas and organ-failure assessment - Why
- Quantify a respiratory or circulatory abnormality requiring urgent support.
- Interpretation and limitations
- A gas helps assess oxygenation, ventilation and acid–base disturbance when clinically indicated. The modified Marshall system formally assesses respiratory, renal and cardiovascular failure; a score of at least 2 in an organ defines failure. Record when the failure began and whether it resolves. Organ failure continuing beyond 48 hours establishes severe acute pancreatitis, irrespective of the height of lipase or an isolated inflammatory marker.
- 04
Inflammatory markers and clinical prediction - Why
- Support risk stratification without substituting for clinical reassessment.
- Interpretation and limitations
- Admission and persistent SIRS, CRP and other validated prediction approaches can identify patients needing closer surveillance. CRP evolves with time, so an early low result is not a reliable exclusion of a severe course. A high CRP does not itself establish infected necrosis or persistent organ failure; interpret it alongside symptoms, physiology and other evidence.
- 05
Imaging prompted by a clinical question - Why
- Identify deterioration, uncertainty or a complication that changes management.
- Interpretation and limitations
- Do not schedule serial CT solely to update a severity label. Cross-sectional imaging is useful for clinical deterioration or intervention planning, and an early scan may be necessary for another urgent diagnosis. Where routine necrosis assessment is needed, later imaging after the initial 72–96 hours is generally more informative. Ultrasound and other tests answer specific causal or competing questions.
04Worked approachesCases with ordered reasoning, an action and a check of the outcome.
01Worked caseEscalation before the duration thresholdA constructed ward example shows early organ support and subsequent classification.+
- 1A 62-year-old man is admitted with confirmed pancreatitis and initially stable circulation. Eight hours later his oxygen requirement has risen from room air to 6 L/min, respiratory rate is 30/min, blood pressure is 88/54 mmHg and urine output has been 12 mL/hour for three hours. The clinician recognises deterioration and calls the senior team and critical care immediately.
- 2The bedside reassessment includes a blood gas, repeat renal profile, review of fluid already received and examination for congestion or another cause. The team arranges higher-level monitoring and circulatory/respiratory support now; it does not wait two days to decide whether he will ultimately meet the severe category. The handover records the onset of each organ abnormality.
- 3After transfer, assessed resuscitation and organ support, blood pressure and urine output improve, but respiratory failure continues beyond 48 hours from its documented onset. The team now classifies the pancreatitis as severe because failure has persisted. This later classification confirms the observed course; it was not the trigger for the earlier lifesaving escalation.
- 4By day five his respiratory support is reducing and creatinine is approaching baseline. The next handover records those measured trends, ongoing nutrition and fluid goals, and which deterioration would prompt renewed investigation. Improvement in one organ is not treated as proof that the whole illness has resolved.
02Admission risk assessmentChoose monitoring from the current stateUse when a new pancreatitis diagnosis reaches the admitting team.+
- 1Record baseline comorbidity and function, the onset of symptoms, vital signs, oxygen requirement, hydration and organ-function tests. Identify any established failure immediately. Escalate patients needing organ support or close observation to an environment able to deliver it; do not use a low prediction score to override that need.
- 2Use SIRS and an appropriate validated prediction approach as additional risk information. Specify the actual findings in the plan, including which measurements will be repeated and by whom. A patient without failure still needs reassessment because early predictive reassurance is imperfect.
- 3Set explicit review triggers such as increasing oxygen, persistent oliguria, falling pressure, confusion, rising creatinine or abdominal change. Communicate them to the nursing and on-call teams so that deterioration causes an active review rather than another routine set of observations.
03A deteriorating patientReassess cause, support and locationUse when symptoms, observations or organ-function trends worsen during the admission.+
- 1Repeat an ABCDE assessment, summon appropriate senior support and treat immediate threats while gathering information. Check whether the apparent change is explained by pain, medication effects, respiratory pathology, hypovolaemia, overload, infection, bleeding or a developing abdominal complication.
- 2Review previous fluid and analgesic interventions against their observed effects. Examine both perfusion and congestion, and obtain targeted tests rather than repeatedly applying a fixed infusion schedule. Persistent hypotension or respiratory deterioration despite initial measures needs critical-care involvement.
- 3Discuss imaging and pancreatic-centre input when a complication or a difficult course is suspected. Record the organ-failure timeline and the outcome of the review. A falling inflammatory marker should not cancel escalation when oxygen need or renal function is getting worse.
05Feedback, follow-up and evidenceReview outcomes, seek feedback and identify what to improve.
- Specify the frequency of observation and clinical review according to instability and care setting. Record oxygen device and delivery, respiratory effort and mentation with the numeric observations, because a supported normal saturation can conceal worsening respiratory function.
- Review urine output and cumulative intake/output alongside examination, weight where practical, renal trends and signs of congestion. A fluid prescription needs a reassessment point and an action if perfusion fails to improve or breathing worsens.
- Keep an organ-failure timeline. Resolution within 48 hours and persistence beyond 48 hours have different classification implications, but every episode of failure requires action at the time it occurs.
- Reassess whether the current ward or critical-care environment remains appropriate. Communicate measured response, nutrition delivery, unresolved diagnostic questions and named review ownership during transfers rather than handing over only a severity score.
06Special situationsVariants, exceptions and circumstances that change the usual approach.
Classification can legitimately change
A patient may begin without organ failure, develop transient failure and later have a different course than first predicted. Updating the severity category is appropriate when new facts emerge. Preserve the chronology so that another clinician can understand whether a duration threshold has actually been met rather than inferring it from the admission date.
Prediction is probabilistic
A prognostic tool summarises a risk pattern across populations. It does not prove that a specific patient has necrosis, infection or future failure. Use the score to inform observation and planning, while direct physiological evidence determines present treatment. Discordance between the score and the bedside assessment is a reason to review the patient, not to force the findings into the score.
Fluid intolerance changes interpretation
Older adults and people with heart, renal or liver disease may tolerate fluid differently from uncomplicated trial participants. Pulmonary findings after resuscitation deserve assessment for overload as well as inflammatory respiratory failure. Persistent oliguria in a congested patient is not a command to continue boluses; combined underperfusion and congestion needs experienced haemodynamic assessment.
Do not diagnose infection from inflammation
Fever, leucocytosis and a high CRP can accompany sterile pancreatic inflammation. Look for a convincing source and the whole clinical course before attributing deterioration to infected necrosis. Conversely, a patient with new sepsis or another infection still needs prompt treatment even if the pancreatic imaging does not yet show a mature collection.
07Common pitfallsFrequent interpretation and management errors.
- 01
Waiting until the 48-hour definition is satisfied before requesting critical-care support; the threshold classifies persistent failure and does not set the timing of resuscitation.
- 02
Calling a patient stable because saturation is unchanged while oxygen delivery has increased substantially; support requirements are part of the observation.
- 03
Treating an elevated CRP, SIRS or a prediction score as direct proof of infected necrosis, without evaluating sterile inflammation and other causes.
- 04
Repeating a fluid intervention because urine output is low without assessing its previous effect, respiratory status, congestion and competing causes of kidney injury.