Sepsis kills approximately 11 million people annually and remains one of medicine's most time-critical diagnoses. The challenge is that it often presents subtly — and by the time it looks obvious, the window for intervention can be closing.
The Case
A 68-year-old woman is brought to the emergency department by her daughter. The presenting complaint, as written in the triage notes, is 'confusion and not feeling well for two days.' Her daughter says she has been more forgetful than usual, has not eaten much, and had a mild fever the previous night that seemed to settle. On examination, she is alert but slightly disoriented to time. Her temperature is 38.1°C. Heart rate is 102. Blood pressure is 108/70. Respiratory rate is 22. She has been incontinent of urine twice today, which is not her baseline.
This is sepsis until proven otherwise. But it doesn't look like the textbook picture. There is no rigors, no crashing blood pressure, no obvious source of infection screaming at you. She is not pale and sweating. She could easily be triaged as a confused elderly patient for further assessment, with the urgency dialled down by the absence of dramatic clinical signs.
Why Early Sepsis Is Missed
This is the core clinical problem with sepsis. The textbook presentation — high fever, rigors, obvious source, florid haemodynamic instability — is the late presentation. By the time the BP is 80/40, you have already lost time that could have been used more effectively. The early presentation is often exactly this: a slightly confused elderly patient, a subtle tachycardia, a borderline temperature, a respiratory rate that is just a bit too fast.
The Sepsis-3 definition, which replaced SIRS criteria in 2016, defines sepsis as life-threatening organ dysfunction caused by a dysregulated host response to infection. This is conceptually important: sepsis is not just infection plus fever. It is infection producing a systemic response that is causing organs to malfunction. In our case, the confusion is not just 'she's 68 and a bit confused' — it is brain that is not functioning normally, and the question is why.
Using qSOFA and Lactate
The qSOFA criteria give a bedside clinical prompt: respiratory rate ≥22, altered mentation, systolic BP ≤100. Our patient meets two of three. That is not a diagnosis, but it is a signal to take seriously. The full SOFA score is more comprehensive and requires investigations — a full blood count, renal function, liver function, coagulation, and critically, a lactate level.
Lactate is where the physiology becomes important. In sepsis, inadequate tissue perfusion — even when the blood pressure is still technically maintained — leads to anaerobic metabolism and lactate production. A lactate above 2 mmol/L suggests tissue hypoperfusion even when haemodynamics appear stable. A lactate above 4 mmol/L defines septic shock in conjunction with vasopressor requirement. This is why blood pressure alone is a misleading guide. A patient can be compensating — maintaining BP through tachycardia and peripheral vasoconstriction — while their tissues are already hypoperfused.
Working the Case
At this stage, the team faces a familiar emergency department dilemma. The patient does not look critically ill yet. Starting broad-spectrum antibiotics early may feel aggressive. Waiting for more confirmation risks losing valuable time. Many missed sepsis cases happen in exactly this grey zone where the presentation is subtle and the instinct to confirm before treating introduces dangerous delay.
Back to the case. Blood cultures are drawn before antibiotics — this is important, but should not delay antibiotics significantly. Urine is sent for microscopy and culture; the history of incontinence and the age and sex of the patient make a UTI the most likely source. A chest X-ray is ordered to exclude pneumonia. Basic bloods are taken, including a venous gas for lactate.
The lactate comes back at 2.8 mmol/L. Creatinine is 142 (her baseline from an old result in the notes is 88). White cell count is 18.4 with a neutrophilia. CRP is 187. Urine dipstick is strongly positive for nitrites and leucocytes. This is sepsis from a urinary source with early acute kidney injury. The management follows the Sepsis-6: high-flow oxygen if needed, blood cultures, IV antibiotics within one hour of recognition, IV fluid bolus, urine output monitoring with a catheter, and serial lactate measurements.
What Case-Based Learning Teaches
What case-based learning teaches you that the textbook cannot is the texture of the decision-making. The textbook gives you sepsis criteria. A case forces you to apply them to a patient who doesn't announce her diagnosis. You have to generate the question 'could this be sepsis?' from a presentation that doesn't obviously look like sepsis. You have to know that confusion in an elderly patient is an early sign of septic encephalopathy. You have to recognise that a heart rate of 102 and a RR of 22 are not normal, even if they don't look dramatic.
The learning objective here is not just knowing the qSOFA criteria — it is developing the clinical vigilance to flag the patient who meets them without realising they do. That vigilance is only built through repeated exposure to cases like this one. Not the textbook version where the diagnosis is named at the top of the page. Cases where you encounter the patient first and have to work out what is happening — which is, of course, exactly what medicine is.
For every hour that appropriate antibiotics are delayed in sepsis, mortality increases. Studies consistently show a 7–10% increase in mortality for each hour of delay after diagnosis. This is why the urgency matters and why the ability to recognise early sepsis — not just obvious sepsis — is one of the highest-value clinical skills a junior doctor can develop. Simulate it. Practice the recognition. Build the instinct before it counts.
Clinical reasoning improves through repeated exposure to real patient scenarios. Explore interactive patient cases on MediKarya to practise this process directly.