Multicentre observational study of adherence to Sepsis Six guidelines in emergency general surgery. BJS 2017; 104: e165-e171.

Published: 25th January 2017

Authors: N. S. Blencowe, S. Strong, J. Blazeby, R. Daniels, C. Peden, J. Lim et al.


Evidence‐based interventions may reduce mortality in surgical patients. This study documented the prevalence of sepsis, adherence to guidelines in its management, and timing of source control in general surgical patients presenting as an emergency.


Patients aged 16 years or more presenting with emergency general surgery problems were identified over a 7‐day period and then screened for sepsis compliance (using the Sepsis Six standards, devised for severe sepsis) and the timing of source control (whether radiological or surgical). Exploratory analyses examined associations between the mode (emergency department or general practitioner) and time of admission, adherence to the sepsis guidelines, and outcomes (complications or death within 30 days).


Of a total of 5067 patients from 97 hospitals across the UK, 911 (18·0 per cent) fulfilled the criteria for sepsis, 165 (3·3 per cent) for severe sepsis and 24 (0·5 per cent) for septic shock. Timely delivery of all Sepsis Six guidelines for patients with severe sepsis was achieved in four patients. For patients with severe sepsis, 17·6–94·5 per cent of individual guidelines within the Sepsis Six were delivered. Oxygen was the criterion most likely to be missed, followed by blood cultures in all sepsis severity categories. Surgery for source control occurred a median of 19·8 (i.q.r. 10·0–35·4) h after diagnosis. Omission of Sepsis Six parameters did not appear to be associated with an increase in morbidity or mortality.


Although sepsis was common in general surgical patients presenting as an emergency, adherence to severe sepsis guidelines was incomplete in the majority. Despite this, no evidence of harm was apparent.

Full text

Your comments


BJS admin

2 years ago

We read this with great interest and are very impressed with the findings highlighted in the study.

However, we have noted a few mathematical errors and wanted to bring these to the author’s attention.

Table 3 lists the diagnoses for patients with sepsis, severe sepsis and septic shock. It states that the diagnosis is missing for 20 patients, out of a total of 911. However, when looking at the figures for sepsis, adding the number for each diagnosis comes to 872. This would seem to suggest that data is missing for a total of 39 patients.

Table 4 shows adherence to the sepsis six guidelines for each category of sepsis severity. For patients with sepsis and adherence to antibiotics administration, it states that 695 patients had the intervention at any time. The time was recorded for 468 patients. This would suggest that for 227 patients, the time was missing. However the table states that the time was missing for only 197 patients. This leaves 30 patients unaccounted for.

Similarly for sepsis and adherence to the oxygen guideline, 137 patients had oxygen delivered at any time, The time was documented for 107, suggesting that for 30 patients the data was missing. However, the table suggests that data was missing for only 28.

These are our observations and we would be grateful if these could be looked into.

Catriona Wright
Khurram Siddique

Northern General Hospital
Herries Rd
S5 7AU

BJS admin

2 years ago

Thank you for forwarding the correspondence you received from Dr Wright and Mr Siddique regarding mathematical errors relating to the number of patients with missing information in our paper ‘Multicentre observational study of adherence to Sepsis Six guidelines in emergency general surgery’, published 25 January 2017. We have checked through all the tables within the paper and can confirm the following minor errors:

For Table 3, the number of patient with missing data should indeed be 39. This error occurred because of a failure to include patients entered into our database with an unknown diagnosis.

For Table 4, the time of administration of antibiotics was indeed missing for 227 patients instead of the 197 stated. Similarly, the timing of administration of oxygen was missing for 30 patients and not 28, as stated.

We apologise for these errors. We do not feel that these detract from the message of this important paper.

Sean Strong
Centre for Surgical Research
School of Social and Community Medicine
Canynge Hall
39 Whatley Road
Bristol, BS8 2PS