Suspected sepsis in adults: recognition, risk stratification and early management

Clinical framing and first assessment

Sepsis is a clinical syndrome caused by a dysregulated response to infection and may progress rapidly. The practical challenge is that early features overlap with many less dangerous illnesses, while some high-risk patients have subtle presentations. Assessment therefore begins with the possibility of infection and the person’s physiological state rather than waiting for a single diagnostic test. The clinician should consider whether infection is likely, whether organ dysfunction or deterioration is present, and whether the person has factors that make standard observations less reassuring.

Initial assessment should be systematic. Record temperature, heart rate, respiratory rate, blood pressure, oxygen saturation and mental state, and consider capillary refill, urine output and skin or peripheral perfusion where relevant. Establish the likely source of infection from history and examination, but avoid letting an apparent source distract from the severity assessment. Immunosuppression, recent surgery, invasive devices, pregnancy or recent pregnancy, frailty and significant comorbidity can modify presentation and risk. Pregnant or recently pregnant people now have a dedicated NICE pathway and should be assessed using the appropriate guideline.

Risk stratification and NEWS2

Current NICE guidance for people aged 16 or over uses structured risk assessment, including NEWS2 in appropriate settings, alongside clinical judgement. NEWS2 is valuable because it standardises physiological deterioration, but it is not a sepsis diagnostic test. A low score does not remove concern when there are high-risk clinical features, and a high score may reflect causes other than infection. The trend in observations and the clinician’s overall concern matter.

Risk assessment should also recognise people in whom NEWS2 may be less reliable or needs contextual interpretation. Chronic hypoxaemia, altered baseline physiology, frailty, medicines affecting heart rate and communication difficulties can complicate scoring. Escalation should follow the highest relevant risk signal from the overall assessment rather than waiting for every criterion to align. Repeated observations are important because movement over time can be more informative than a single snapshot.

Investigations and identifying the source

Investigations should support two goals: assess organ dysfunction and identify the likely pathogen or source without delaying time-critical treatment in high-risk patients. Depending on severity and setting, this may include blood tests such as full blood count, renal and liver profile, clotting, glucose, C-reactive protein and lactate, together with blood cultures before antimicrobials when this can be achieved without clinically significant delay. Additional microbiology, urine testing and imaging are guided by the suspected focus.

Lactate can help identify tissue hypoperfusion and risk but is not specific for sepsis. A normal value should be interpreted with the rest of the clinical picture. Imaging may identify pneumonia, obstruction, collection or another focus requiring intervention. Source control is a core management principle: an abscess, infected device, obstructed urinary tract or surgical source may not improve with antimicrobials alone. Early specialist or surgical involvement is appropriate when source control may be needed.

Antimicrobial treatment and timing

The timing of antibiotics should reflect the person’s risk category and the likelihood of bacterial infection. NICE moved away from a simplistic approach in which every person with possible sepsis receives immediate broad-spectrum antibiotics irrespective of risk. High-risk patients need rapid antimicrobial treatment in accordance with the guideline and local antimicrobial policy, while lower-risk patients may benefit from focused assessment that reduces avoidable antibiotic exposure. The key is not to create delay in a patient who is deteriorating or has high-risk features.

Before treatment, document suspected source, allergy history, relevant recent microbiology, recent antimicrobial exposure and factors affecting dose or drug choice such as renal function. Local antimicrobial guidance should determine the empirical regimen because resistance patterns and formulary choices vary. Review therapy when microbiology and clinical response become available, narrowing, stopping or changing treatment when appropriate. This stewardship step is part of safe sepsis care rather than an optional later exercise.

Fluids, oxygen and haemodynamic support

Fluid resuscitation is individualised. Patients with hypotension or evidence of hypoperfusion may need intravenous crystalloid, but age, heart failure, renal impairment and the risk of fluid overload influence volume and reassessment. Give fluid in measured steps with repeated evaluation of blood pressure, perfusion, respiratory status and signs of congestion rather than assuming that a fixed total volume is suitable for every adult. Persistent haemodynamic instability requires senior review and consideration of critical care support.

Oxygen should be titrated to an appropriate saturation target rather than given without regard to chronic respiratory disease. Correct hypoglycaemia and other immediately reversible abnormalities. Urine output and mental state are useful markers of response. When shock persists despite initial resuscitation, vasopressor support and invasive monitoring may be needed in a critical care environment. Early escalation is preferable to repeated ward-level interventions without improvement.

Reassessment, escalation and communication

Sepsis care is a repeated assessment process. Review observations, consciousness, perfusion, urine output, lactate where indicated, response to fluids and the evolving source diagnosis. A patient initially considered lower risk may cross an escalation threshold as physiology changes. Conversely, improvement and a clear alternative diagnosis may justify de-escalation. The plan should state when observations will be repeated and which changes require senior or critical care review.

At handover, communicate the suspected source, risk category, antimicrobial timing, cultures obtained, fluids given, relevant results and outstanding actions. If sepsis is not diagnosed and the person is discharged, provide safety-netting on symptoms that should prompt urgent reassessment. Clear documentation is especially important when antibiotics are deferred because the reason, review plan and escalation criteria should be visible to the next clinician.

Practice points and limitations

Common failure modes include treating NEWS2 as a diagnostic result, delaying escalation while awaiting tests, giving antimicrobials without considering source or local policy, failing to reassess after fluids, and overlooking the need for source control. Equally, broad-spectrum antibiotic use in low-risk illness without adequate assessment creates avoidable harm. The current NICE approach supports urgency where risk is high and more discriminating assessment where it is safe to do so.

This resource applies to people aged 16 or over who are not pregnant or recently pregnant and summarises NICE NG253 at a professional-reference level. Local sepsis pathways, antimicrobial policies and critical care escalation criteria remain essential. Children and young people under 16 and people who are pregnant or recently pregnant have separate NICE guidance.

Diagnostic uncertainty, stewardship and post-acute review

Sepsis pathways operate under uncertainty. Early in the presentation, the clinician may know that the person is physiologically unwell before the source or even the infectious nature of the illness is clear. The safest approach is to state the working diagnosis and level of concern explicitly, then update it as data arrive. Differential diagnoses such as pulmonary embolism, pancreatitis, adrenal crisis, toxicological illness, haemorrhage and cardiogenic shock can produce overlapping physiology and may need parallel investigation.

Antimicrobial stewardship should continue after the first dose. Review cultures, imaging, inflammatory markers and the clinical response at the earliest useful opportunity. If bacterial infection becomes unlikely, unnecessary antibiotics should be stopped according to local policy. If an organism or source is identified, narrow therapy when appropriate and ensure that duration is linked to the diagnosis rather than inherited from a generic sepsis label. In patients with recurrent infection, recent hospital exposure or previous resistant organisms, microbiology advice may materially change empirical and definitive treatment.

Recovery from sepsis can extend beyond resolution of the acute physiological disturbance. At discharge or transfer, ensure that outstanding microbiology, imaging or source-control actions have an identified owner. Patients and families may need information about expected recovery, medication changes and warning symptoms. Where the episode revealed a new underlying problem, such as obstructive uropathy, immunosuppression or poorly controlled diabetes, follow-up should address that driver rather than closing the episode when observations normalise.

Older adults and people with immunosuppression deserve particular attention because fever and inflammatory responses may be muted. New confusion, reduced mobility, poor intake or functional decline can be prominent features. Medication history may also alter physiology, for example beta blockers limiting tachycardia. The absence of a classic septic picture should therefore be weighed against baseline function, trajectory and vulnerability. In these groups, early senior review can be appropriate even when individual observations appear only moderately abnormal.