What sepsis is, and why the clock is the whole problem

Sepsis is one of the most common ways people die of an infection, and one of the most widely misunderstood. It is not a specific germ and not a synonym for a bloodstream infection. Under the current international definition, agreed in 2016 and known as Sepsis-3, sepsis is "life-threatening organ dysfunction caused by a dysregulated host response to infection" [s1]. The damage that kills is done largely by the body's own reaction — the immune and clotting systems, mobilised against a pathogen, start injuring the lungs, kidneys, and circulation of the person they are meant to protect.

That reframing matters because it moves the danger from the bug to the response. An infection that begins as a chest infection, a urinary infection, or an infected wound becomes sepsis when the reaction to it starts shutting organs down. Septic shock, the most severe form, is defined in Sepsis-3 as sepsis with circulatory and metabolic abnormalities severe enough that in-hospital mortality exceeds 40% — specifically, a patient needing drugs to maintain blood pressure and with a raised blood lactate despite adequate fluid [s1].

The scale is easy to underestimate

Sepsis rarely appears on a death certificate as the headline cause, which is part of why its scale was long undercounted. A Global Burden of Disease analysis published in The Lancet in 2020 tried to measure it directly and estimated 48.9 million cases of sepsis worldwide in 2017 and 11.0 million sepsis-related deaths — about 19.7% of all deaths globally that year [s2]. The same analysis found the burden falls hardest on the poorest regions and on the young: a large share of cases were in children, and age-standardised incidence was highest in sub-Saharan Africa [s2]. These are modelled estimates with wide uncertainty, and the authors are explicit about the limits of the data feeding them [s2]. But even the lower bounds describe a leading global cause of death.

Why early recognition is the whole game

The reason clinicians treat suspected sepsis as an emergency is that the evidence keeps returning to time. The most-cited demonstration is a 2006 study of patients in septic shock, which found that among those who survived, effective antimicrobial treatment was started far sooner than among those who died — and that each hour of delay in giving an effective antibiotic over the first six hours after the onset of low blood pressure was associated with a measurable fall in survival [s4]. The study was observational, so it cannot prove the delay itself caused every death, and later work has debated the exact shape of the curve. But the direction has held up across many datasets, and it is why guidelines are built around speed.

The Surviving Sepsis Campaign, the main international guideline, recommends that for adults with possible septic shock or a high likelihood of sepsis, antimicrobials be given immediately, ideally within one hour of recognition [s3]. It also recommends measuring blood lactate, drawing blood cultures before antibiotics where that does not delay them, and giving intravenous fluid for sepsis-induced low perfusion [s3]. The guideline grades many of these recommendations on low-quality evidence and is candid that the one-hour target for less clear-cut sepsis, as opposed to outright shock, is contested [s3]. The underlying logic is not: a dysregulated response left running does more organ damage the longer it runs.

What recognition actually looks like

Because sepsis has no single sign, the definition leans on organ dysfunction rather than the infection itself. Sepsis-3 formalised this with the Sequential Organ Failure Assessment (SOFA) score, where a rise of two or more points in the setting of infection marks organ dysfunction [s1]. For rapid bedside use outside intensive care, the same group proposed a simpler screen, qSOFA, flagging patients with at least two of three findings: a fast respiratory rate (22 breaths a minute or more), altered mental state, or low systolic blood pressure (100 mmHg or below) [s1]. qSOFA was designed to prompt closer assessment, not to diagnose, and later studies found it too insensitive to use as a sole screening tool — a limitation the wider field now acknowledges.

For a non-clinician the practical content is narrower and is about escalation, not diagnosis: an infection that is getting rapidly worse rather than better, especially with confusion, very fast breathing, a drop in urine, mottled or clammy skin, or a general sense of severe illness, is the pattern that warrants urgent medical assessment. None of that identifies sepsis on its own, and this article does not attempt to. What the evidence supports is narrower and firmer: sepsis is the body's response to infection turning against its own organs, it is common and lethal, and the single lever that most consistently changes outcomes is how quickly it is recognised and treated [s3][s4].