+ Direct Access
August 7, 2017


Sepsis-associated encephalopathy (SAE) was defined by neurologic symptoms consistent with delirium and Glasgow Coma Scale (GCS) lower than 15. Excessive and prolonged neuro-inflammation, alterations of the blood–brain barrier and microglial activation were supposed to cause the macroscopic brain tissue damage observed with neuro-imaging and autopsy in patients with SAE. Data on potentially modifiable contributing factors (severe hypoxia, systemic hypotension, metabolic disturbances, use of sedative agents, antibiotics and steroids) to SAE are scarce.

Sonneville et al. [1] conducted a retrospective analysis of the prospective multicentre OUTCOMEREA database with the aim to identify the determinants and prognostic value of SAE in critically ill patients at ICU admission. Among 18713 adult patients admitted in 12 French ICUs between 1997 and 2014, 2513 patients with severe sepsis or septic shock at ICU admission were included in the study. Sepsis had mainly a pulmonary and intra-abdominal origin. Bacteraemia was diagnosed in a quarter of the patients. On the day of ICU admission, 2208 (88%) patients had septic shock and 1071 (43%) patients required invasive mechanical ventilation.

SAE was observed in 1341 (53%) patients, median GCS was 11 (4–14). Among these patients, 19% had a GCS of 15 and features of delirium, 23% had GCS of 13–14, 18% had a GCS of 9–12, and 40% had a GCS of 3–8. Patients with SAE most often presented hypoactive delirium. SAE were predominantly diagnosed in older patients and in those who had a history of chronic psychiatric, neurological and cognitive impairment, long-term use of psychoactive drugs, alcohol abuse, chronic liver disease, and immune-depression. SAE was associated with ICU admission diagnosis, SAPS 2 and SOFA scores, multi-organ (respiratory, liver and renal) failure, use of sedatives (propofol, midazolam, opioids) and steroids, hypercapnia, hypo- and hyper-glycaemia, hypernatremia, infection source, bacteraemia, S. aureus infection and septic shock. By contrast, PaO2 was comparable between SAE and non-SAE patients. GCS of 13–14 at ICU admission was associated with increase in mortality rate.

In this retrospective analysis of a large prospective multicentre database, Sonneville et al. [1] observed that 50% of the patients suffered from SAE at ICU admission. Multi-organ failure, medications (antibiotics, sedatives steroids) and metabolic alterations were likely to play a role in the pathophysiology of SAE due to previously observed neurologic side effects.

The association between S. Aureus and SAE is multifactorial: α-haemolysin was identified as neurotoxic, S. Aureus bacteraemia was associated with arterial thromboembolic events and this bacterium is frequently isolated in patients with infective endocarditis.

Progress in critical care and sepsis management over the past 20 years has led to a dramatic decrease of sepsis mortality, although the outcome of comatose septic patients at admission has not significantly decreased and even mild alteration in mental status at ICU admission (GCS 13–14) was independently associated with mortality. The impact of transient (1 day) or persistent (more than 1 day) SAE as well as SAE developing during ICU stay and long term follow-up remain to be investigated.

Article review prepared and submitted by Salvatore Lucio Cutuli (EJRC member) and Gennaro de Pascale (NEXT committee member), Department of Anesthesiology and Intensive Care, Università Cattolica del Sacro Cuore, Fondazione Policlinico Agostino Gemelli, Rome, Italy. 


  1. Sonneville R, de Montmollin E, Poujade J, Garrouste-Orgeas M, Souweine B, Darmon M, Mariotte E, Argaud L, Barbier F, Goldgran-Toledano D, Marcotte G, Dumenil AS, Jamali S, Lacave G, Ruckly S, Mourvillier B, Timsit JF. Potentially modifiable factors contributing to sepsis-associated encephalopathy. Intensive Care Med. 2017 Aug;43(8):1075-1084.

Comment on this news