Electrocardiographic abnormalities are common in sepsis but are generally nonspecific and should not be used alone to diagnose sepsis. Findings may reflect systemic inflammation, myocardial injury, autonomic disturbance, metabolic abnormalities, or other acute illness.

Common findings include:

• Sinus tachycardia is frequent in sepsis. In a large emergency department cohort, sinus tachycardia was not independently associated with poor outcome after adjustment for other factors.^[1]
• QRS widening and bundle branch block may occur but are nonspecific and are not diagnostic of sepsis.^[1]
• QT prolongation is common. In a large emergency department sepsis cohort, automated interpretation identified QT prolongation in 54.4% of patients. New-onset QT prolongation occurred in approximately 23% of septic patients and was independently associated with mortality.^[1][2]
• ST-segment elevation may occur and can mimic acute coronary syndrome. In the emergency department cohort, automated interpretation flagged ST elevation in 10.4%, whereas only 1.7% met cardiologist-confirmed acute coronary syndrome criteria.^[1]
• T-wave abnormalities may occur. Abnormal T waves were present in 8.9% of patients in the emergency department cohort and were associated with worse outcomes (OR 1.34, 95% CI 1.01–1.76).^[1]
• Atrial fibrillation/flutter is common in severe sepsis. Reported incidence is approximately 6%–22% in severe sepsis, with a pooled new-onset atrial fibrillation prevalence of approximately 13% overall (about 7% on general wards to 21% in the ICU). In a cohort of critically ill patients with sepsis, cumulative rates of new-onset atrial fibrillation were approximately 10% with sepsis, 22% with severe sepsis, and 40% with septic shock.^[3][4][5]

New-onset atrial fibrillation in sepsis is associated with increased stroke risk (pooled OR approximately 1.17) and recurrent atrial fibrillation. However, routine acute anticoagulation is not favored, and CHA₂DS₂-VASc performs poorly for stroke prediction in this setting. Anticoagulation decisions should be individualized; detailed management is addressed in the medical therapy microchapter.^[4][3]

ST-segment elevation in sepsis does not necessarily indicate acute coronary occlusion. Critical illness may produce ECG patterns that mimic acute myocardial infarction, and ECG abnormalities should be interpreted together with the clinical presentation and cardiac biomarkers.

In a cohort of critically ill patients with ECG patterns interpreted as ST-segment elevation myocardial infarction, most patients did not have a corresponding troponin rise, supporting caution in attributing ST-segment elevation to type 1 myocardial infarction.^[6]

Elevated troponin levels are common in sepsis and are associated with short-term mortality, but troponin elevation alone does not establish the mechanism of myocardial injury or a diagnosis of type 1 myocardial infarction.^[7]

QT prolongation is a clinically relevant ECG abnormality in sepsis because new-onset QT prolongation has been associated with mortality. ECG interpretation should account for potentially reversible contributors to QT prolongation, including metabolic abnormalities and medications.^[2]

Fever can unmask a Brugada ECG pattern. Recognition of a Brugada pattern during febrile illness is important because fever may precipitate ventricular arrhythmias in susceptible patients. Fever should be promptly treated, while specialist evaluation and longer-term management are addressed outside this microchapter.^[8]