Most UK anaesthetists would employ epidural analgesia for an emergency laparotomy, according to a national survey published in Anaesthesia [1]. There appears to be growing concern over the use of this technique in patients with signs of sepsis. However, there is little or no evidence of the benefits and safety or otherwise of performing central neuraxial blocks for postoperative analgesia in patients with sepsis [2]. The development of policies for dealing with this clinical situation may therefore be best informed by current opinion. We have attempted to define this opinion by conducting a survey of lead clinicians of acute pain teams in the UK.

We sent a questionnaire addressed to the lead clinician of the acute pain team to each of the 304 hospitals with an anaesthetic department in the UK. If the questionnaire was not returned within 8 weeks, a repeat questionnaire was sent. The questionnaire asked whether the hospital had a policy for the use of epidural analgesia in septic patients, and sought the opinion of the respondent concerning epidural analgesia in two hypothetical patients with diagnoses of bowel perforation, requiring laparotomy. The first patient was thought to have a localised bowel perforation, and was apyrexial, with a white cell count (WCC) of 12 × 109.l−1, and looked well. The second patient had a diagnosis of a perforated diverticulum, looked ‘sick’ and met the criteria for systemic inflammatory response syndrome (SIRS), with a temperature of 39 °C and WCC of 15 × 109.l−1, but was haemodynamically stable and had normal clotting studies.

We then examined whether the decision would be influenced by the availability of a higher care facility (level 2 or above). We also examined where the clinician had decided against epidural analgesia and whether this decision was influenced by three specific factors: risk of developing an epidural abscess; haemodynamic instability; and contraindication to activated protein C. Respondents were also able to state other reasons for the decision.

After following up initial non-responders, we received 211 replies, a 69% return rate. Of these, 205 had an acute pain service and went on to complete the questionnaire. All employed epidural analgesia as a method of postoperative analgesia, but only 185 (90%) on a regular basis for laparotomy. Only five hospitals (2%) had guidelines for use of epidural analgesia in the presence of sepsis.

One hundred and fifty-two (82%) of the 185 respondents who routinely used epidural analgesia for laparotomies stated that they would do so for the patient with a suspected bowel perforation, but no signs of SIRS, if a high dependency unit (HDU) bed was available. This dropped to 110 (59%) if no HDU bed was available. For the patient with SIRS, 49 (27%) respondents would use epidural analgesia if a HDU bed was available, and 13 (7%) if not.

Of those who would not use epidural analgesia, 116 cited the danger of epidural abscess as a reason for not doing so, 102 cited haemodynamic instability, 23 the contraindication to activated protein C, and 14 gave other reasons (multiple reasons were allowed), most commonly the potential for development of coagulopathy.

Whilst this survey has not demonstrated a consensus, it is clear that a majority of clinicians with responsibility for acute pain services consider the use of epidural analgesia appropriate for patients with peritonitis who do not show signs of systemic sepsis at the time of surgery. In patients with SIRS, a majority would avoid epidural analgesia, the decision being influenced, to some extent, by the availability of a HDU bed, although many respondents made the point that the decision whether to care for the patient in a ward or higher care facility was more related to the patient’s overall condition and co-morbidities than to epidural analgesia.

The risk of epidural abscess was the most common concern expressed. The effect of systemic sepsis on this risk is not known, although it may be more influenced by the duration of epidural analgesia and patients’ impaired immune status. The most common pathogen causing epidural abscesses related to central neuraxial block is Staphylococcus aureus [3], suggesting perhaps that the risk related to gram-negative sepsis may be relatively small. We received many comments to the effect that the presence of respiratory compromise may sway the decision towards epidural analgesia, presumably due to the belief that it may improve outcome in this group of patients. This serves as a reminder that a risk–benefit analysis needs to be performed for each patient. We believe that guidelines are required, and we trust that in the absence of hard evidence, this snapshot of current opinion is informative.

No external funding and no competing interests declared.


  1. Top of page
  2. References