Pneumonia and Sepsis

TL;DR. Pneumonia is infection of the air sacs at the far end of the lungs. They fill with fluid and pus, so oxygen cannot cross into the blood, and the patient drowns slowly in their own inflammatory response. Sepsis is what happens when the body's reaction to any infection stops being local and becomes a systemic emergency: blood vessels leak and dilate, clots form in small vessels, blood pressure falls, and organs fail one after another. Sepsis is not a specific germ; it is a response, and it can follow a chest infection, a urinary infection, a burst appendix, or an infected cut. Together they are among the largest causes of death on earth, and both are highly time-dependent: outcome depends on how quickly treatment starts.

Key takeaways

• Sepsis accounts for roughly 11 million deaths a year, close to 20 percent of all deaths worldwide, and about 85 percent of cases occur in low- and middle-income countries.
• Pneumonia is the leading infectious cause of death in children under 5, killing hundreds of thousands a year, almost all preventable or treatable.
• Sepsis is defined by organ dysfunction, not by fever or a positive blood culture. Many patients with sepsis never have a germ identified.
• Time matters enormously. In septic shock, each hour of delay in effective antibiotics is associated with measurably higher mortality.
• The classic warning signs are non-specific, which is why sepsis is missed: confusion, fast breathing, a very high or very low temperature, and simply looking severely unwell.
• Vaccination prevents a large share of it: pneumococcal, Hib, influenza, COVID-19, and measles vaccines all reduce pneumonia and its downstream sepsis.

What they are

In short: Pneumonia is infection filling the air sacs; sepsis is the body's response to any infection turning systemic and damaging organs.

Pneumonia is infection of the lung parenchyma: the alveoli, the tiny sacs where oxygen crosses into blood. They fill with fluid, bacteria, and immune cells, a state radiologists call consolidation, and that region of lung continues to receive blood while contributing no oxygen. The result is low blood oxygen and a body working harder to breathe.

Categories matter because they predict the organism:

Sepsis is defined (Sepsis-3, 2016) as life-threatening organ dysfunction caused by a dysregulated host response to infection. Septic shock is sepsis with circulatory and metabolic failure severe enough that blood pressure cannot be maintained without drugs, and it carries mortality above 40 percent.

Don't be confused: sepsis is not "blood poisoning" and does not require bacteria in the blood. Bacteraemia means organisms in the bloodstream and can be transient and harmless. Sepsis means the response has become damaging. A patient can be severely septic with sterile blood cultures, and can have bacteraemia without sepsis. The dangerous version of the old term is the belief that if cultures are negative, the patient is not that ill.

The history

In short: Osler called pneumonia the captain of the men of death, and sepsis only acquired a usable definition in 2016.

Pneumonia was the archetypal deadly infection of the pre-antibiotic era. William Osler called it "the captain of the men of death," borrowing from Bunyan, and noted with some resignation that it often ended the suffering of the old and infirm quickly.

The pneumococcus was identified in 1881, independently by Pasteur and Sternberg. Serum therapy in the 1910s to 1930s helped modestly. Sulfonamides in the late 1930s and penicillin in the 1940s transformed prognosis: mortality from pneumococcal pneumonia in young adults fell from roughly 30 percent to under 10 percent.

Sepsis took much longer to define usefully. Definitions in 1991 and 2001 relied on "systemic inflammatory response syndrome" criteria (heart rate, temperature, respiratory rate, white cell count) that were so unspecific that they captured almost anyone with flu. The 2016 Sepsis-3 definition moved the centre of gravity to measurable organ dysfunction, which improved both research and bedside recognition.

Two more strands shaped modern care. Early goal-directed therapy (Rivers, 2001) proposed an aggressive protocol of fluids and monitoring and appeared to halve mortality; three large multi-centre trials a decade later found no benefit over usual care, by which time usual care had improved partly because of the original study. And vaccination changed the epidemiology: conjugate vaccines against Haemophilus influenzae type b and against the pneumococcus removed leading causes of childhood pneumonia and meningitis in countries that adopted them.

What actually goes wrong

In short: In pneumonia the air sacs fill so oxygen cannot cross; in sepsis the vessels leak and dilate everywhere at once and organs fail in sequence.

In pneumonia

Organisms reach the alveoli by inhalation, by aspiration of mouth contents, or through the blood. Alveolar macrophages usually clear small inocula; when they cannot, they release cytokines that recruit neutrophils. Capillaries leak, and the air sacs fill with fluid, neutrophils, and debris.

Two consequences follow. Gas exchange fails locally while blood keeps flowing past the consolidated area, so deoxygenated blood mixes back into the arterial circulation (shunt), which is why oxygen alone sometimes fails to correct the saturation. And the lung becomes stiffer, so breathing takes more work, which is why fast, shallow breathing is such a reliable sign.

In sepsis

The same defence, unrestrained and systemic:

1. Recognition: pathogen molecules trigger toll-like receptors on immune cells everywhere, not just at the infection site.
2. Cytokine flood: TNF-alpha, interleukin-1, interleukin-6, and others circulate in quantity.
3. Vasodilation: nitric oxide production surges, arteries dilate, and blood pressure falls. The patient can be warm and flushed rather than cold and pale, which misleads people expecting classic shock.
4. Capillary leak: the endothelium becomes permeable and fluid moves out of vessels into tissue. The circulation is simultaneously overloaded with total body water and short of circulating volume.
5. Coagulation activation: clotting is switched on throughout the microcirculation while clotting factors and platelets are consumed, producing both microthrombi and bleeding. In its extreme form this is disseminated intravascular coagulation.
6. Cellular failure: even where oxygen arrives, mitochondria use it poorly.
7. Organ dysfunction: kidneys (falling urine output), lungs (ARDS), liver (jaundice), brain (confusion, often the first sign in the elderly), heart, and clotting system.

Later in the course, many patients swing into an immunosuppressed phase, vulnerable to secondary infections. This two-phase pattern is one reason simple anti-inflammatory strategies have repeatedly failed in sepsis trials: the right intervention depends on which phase the patient is in, and we cannot yet measure that reliably at the bedside.

What it does to the body

Pneumonia: cough (productive or dry), fever, breathlessness, pleuritic chest pain (sharp, worse on inspiration), fatigue. In older adults it frequently presents without fever or cough, as confusion, falls, or simply "off legs," which delays diagnosis. Complications include pleural effusion, empyema (pus in the pleural space, requiring drainage), lung abscess, respiratory failure, and sepsis.

Sepsis: fever or hypothermia, fast heart rate, fast breathing, low blood pressure, confusion, reduced urine output, mottled skin, and a characteristic sense among experienced clinicians that the patient simply looks gravely ill. In septic shock, mortality is high even with optimal care.

Meningococcal sepsis deserves specific mention because of its speed: it can progress from feeling unwell to death within hours in a previously healthy child or young adult, and the non-blanching rash (which does not fade under pressure) appears late. Waiting for the rash is a documented cause of preventable death.

Is it deadly?

• Sepsis: an estimated 48.9 million cases and 11 million deaths a year, about 19.7 percent of all global deaths, with the burden falling heavily on low- and middle-income countries and on children.
• Pneumonia: the leading infectious cause of death in children under 5, causing hundreds of thousands of deaths annually, nearly all in low- and middle-income countries; it is also a leading cause of death in adults over 65 everywhere.
• Mortality by severity: community-acquired pneumonia treated at home has mortality under 1 percent; hospitalised, roughly 5 to 15 percent; requiring intensive care, 20 to 50 percent. Septic shock exceeds 40 percent.
• Pneumonia is frequently the final event in frail patients with other advanced disease, which is what Osler was describing, and that remains a legitimate context for choosing comfort-focused care.

Is it contagious?

The infections that cause pneumonia are often contagious. Sepsis itself is not.

Pneumococcus, Haemophilus, Mycoplasma, influenza, RSV, COVID-19, and tuberculosis all spread person to person by the respiratory route. Many people carry pneumococcus in their nose harmlessly, and disease occurs when it descends into the lungs in a susceptible person, often after a viral infection has damaged the airway lining.

Some causes are not contagious at all: Legionella comes from contaminated water systems (cooling towers, hot water systems, spa pools) and is inhaled as aerosol, with no person-to-person spread. Aspiration pneumonia comes from the patient's own mouth flora.

Sepsis is a response, so a person in septic shock cannot give sepsis to anyone. They may carry a transmissible organism, which is a separate question.

Who gets it

Age at both extremes. Infants have immature immunity and small airways; older adults have weaker cough, reduced immunity, and more comorbidity.

Risk factors: smoking (which paralyses the ciliary escalator), COPD and asthma, diabetes, heart failure, chronic kidney and liver disease, alcohol use, immunosuppression (HIV, chemotherapy, steroids, biologic drugs), malnutrition, and impaired swallowing after stroke or in dementia.

Environmental: household air pollution from cooking with solid fuels is a major contributor to childhood pneumonia in low-income countries, alongside crowding and undernutrition.

Post-surgical, post-injury, and hospital patients are at high sepsis risk from urinary catheters, intravenous lines, ventilators, and surgical sites, which is why hospital infection control is a life-safety system rather than housekeeping.

Sepsis is also an equity problem. Maternal sepsis remains a leading cause of maternal death worldwide, and neonatal sepsis a leading cause of newborn death, both concentrated where clean delivery, antibiotics, and oxygen are scarce.

Treatment, and how it works

In short: Antibiotics within the hour, careful fluids, and above all source control, because no antibiotic sterilises undrained pus.

Pneumonia

• Antibiotics, chosen empirically by setting and severity, then narrowed when culture results arrive. Community-acquired pneumonia is usually treated with amoxicillin or a macrolide, or a combination in more severe cases, adjusted to local resistance patterns. Viral pneumonia does not respond, but distinguishing viral from bacterial at the bedside is genuinely hard, which is a major driver of antibiotic overuse (Chapter 36).
• Oxygen, targeted to a saturation range rather than as much as possible, since too much oxygen also causes harm.
• Fluids, cautiously.
• Drainage of an empyema, because antibiotics do not sterilise a collection of pus.
• Ventilatory support where needed, from high-flow nasal oxygen and non-invasive ventilation through to intubation.
• Chest physiotherapy and early mobilisation in recovery.

Sepsis

The core is simple and time-critical, usually organised as a bundle to be completed within the first hour of recognition:

1. Measure lactate, a marker of tissue hypoperfusion, and repeat it to track response.
2. Take blood cultures before antibiotics, if this does not delay them.
3. Give broad-spectrum antibiotics immediately. In septic shock, mortality rises with each hour of delay.
4. Give intravenous fluid for hypotension or high lactate, then reassess rather than continuing indefinitely, because too much fluid worsens outcomes through oedema.
5. Start vasopressors (usually noradrenaline) if pressure remains low despite fluid. These constrict dilated vessels to restore perfusion pressure.
6. Control the source. This is the step most often forgotten and most often decisive: drain the abscess, remove the infected line, relieve the obstructed kidney, operate on the perforated bowel. Antibiotics cannot sterilise undrained pus.

Additional measures: corticosteroids in shock requiring ongoing vasopressors, kidney replacement therapy for failing kidneys, lung-protective ventilation with small tidal volumes for ARDS (one of the few interventions that clearly reduces mortality in critical care), and glucose control.

Screening tools such as NEWS2 and qSOFA help identify deteriorating patients early, and electronic alerts in hospital records are increasingly used, though their real-world benefit depends heavily on whether anyone responds to them.

What treatment costs

• Broad-spectrum antibiotics save lives in sepsis and drive resistance, damage the gut microbiome, and cause C. difficile colitis. The resolution is not to give them reluctantly in sepsis; it is to de-escalate to a narrow agent within 48 hours once the organism is known, and to stop when the course is done.
• Fluids: too little leaves organs underperfused, too much causes pulmonary oedema and worse outcomes. Modern practice has moved from "fill them up" to careful, reassessed boluses.
• Vasopressors: restore blood pressure at the cost of reduced perfusion in fingers, toes, and gut in extreme doses.
• Mechanical ventilation: sedation, delirium, muscle wasting, ventilator-associated pneumonia, and lung injury from the ventilator itself if volumes are too large.
• Oxygen: liberal oxygen targets have been associated with worse outcomes in several trials, so more is not better.

What the person can do

In short: Vaccinate, stop smoking, look after teeth, and know that asking a clinician directly whether this could be sepsis is a legitimate question.

• Vaccinate. Pneumococcal vaccination for older adults and high-risk groups, Hib and pneumococcal conjugate vaccines in childhood, annual influenza vaccine, COVID-19 vaccine, and RSV vaccine where offered. Measles vaccination prevents a great deal of secondary pneumonia in children.
• Stop smoking, which restores mucociliary clearance over months.
• Look after teeth and gums, especially with swallowing difficulty. Oral hygiene programmes measurably reduce aspiration pneumonia in nursing homes and hospitals.
• Get help early with warning signs: breathing fast, breathlessness at rest, confusion, a fever that will not settle, or feeling dramatically worse than a normal illness.
• Know the sepsis question. If you or someone you care for is very unwell with a possible infection, asking a clinician directly "could this be sepsis?" is a legitimate, useful question that has been promoted by patient safety campaigns because it works.
• After discharge, expect a long recovery. Fatigue for weeks after pneumonia is normal; a cough can persist for a month.

Living with it

Post-sepsis syndrome and post-intensive care syndrome are the underappreciated sequel. Survivors of severe sepsis frequently have persistent muscle weakness, fatigue, cognitive impairment, anxiety, depression, and post-traumatic stress, for months or permanently. Rehospitalisation in the following year is common, and mortality remains elevated for years after apparent recovery. Families also experience high rates of psychological distress after an ICU admission.

This matters because sepsis care has historically ended at hospital discharge. Structured follow-up, rehabilitation, and medication review after critical illness are increasingly recognised as part of treatment rather than optional aftercare.

What's next

• Rapid diagnostics. Tests that identify the organism and its resistance profile in hours instead of days would allow narrow-spectrum therapy from the start.
• Host-response biomarkers, distinguishing bacterial from viral infection and identifying which sepsis phase a patient is in, which is the prerequisite for any immune-modulating therapy to work.
• Machine learning early warning systems built into hospital records, with the honest caveat that early deployments have had mixed real-world results.
• Pneumococcal vaccines with broader serotype coverage, since vaccinating against some types allows others to take their place (serotype replacement).
• Oxygen access, which the COVID-19 pandemic revealed as a critical gap: medical oxygen is not reliably available in many hospitals worldwide, and it is one of the cheapest life-saving interventions in this book.

Sources and notes

Sepsis global burden: Rudd et al., The Lancet, 2020 (48.9 million cases, 11 million deaths, 19.7 percent of global deaths in 2017). Sepsis-3 definitions: Singer et al., JAMA, 2016. Antibiotic timing in septic shock: Kumar et al., Critical Care Medicine, 2006, and subsequent analyses; the strength of the hour-by-hour relationship is debated but the direction is not. Early goal-directed therapy: Rivers et al., NEJM, 2001, and the ProCESS, ARISE, and ProMISe trials, 2014 to 2015. Lung-protective ventilation: ARDS Network, NEJM, 2000. Childhood pneumonia mortality: WHO and UNICEF estimates. Surviving Sepsis Campaign guidelines, most recent edition, for bundle content. Post-sepsis outcomes: Prescott and Angus, JAMA, 2018.

Open questions. No immune-modulating therapy has succeeded in sepsis despite decades of trials, probably because sepsis is not one biological state. Optimal fluid volumes and blood pressure targets remain actively debated. Whether early warning algorithms improve outcomes in practice is unresolved.

Next: the infection that has killed more people than any other in history, and is still killing over a million a year. 👉