Page 235 Acute Pain Management

The incidence of transient neuropathy after epidural analgesia in large case series was in the
range of 0.013 to 0.023% (Xie & Liu, 1991 Level IV; Tanaka et al, 1993 Level IV; Auroy et al, 1997
Level IV).
Epidural haematoma
A major concern is the development of an epidural haematoma with subsequent, potentially
permanent, spinal cord injury. A review including case series involving over 1 335 000 patients
with epidural analgesia reported seven cases of haematoma (0.0005%) (Wulf, 1996 Level IV). On
the basis of this case series the possible incidence is in the order of 1 in 100 000 at the upper
limit of the 95% confidence interval. The Swedish case series quoted above puts the overall
risk of epidural haematoma after epidural blockade at 1 in 10 300 (Moen et al, 2004 Level IV). An
even higher incidence of epidural haematoma (1:3100) has been estimated for epidural
analgesia in association with inappropriate low molecular weight heparin (LMWH) dose
regimens (Horlocker et al, 2003) (see Section 7.4).
A review of data from publications reporting adverse events after obstetric epidural analgesia
reported a risk estimate of 1:168 000 for epidural haematoma (Ruppen et al, 2006a Level IV).
A meta‐analysis of the risks of epidural haematoma associated with epidural anaesthesia/
analgesia in cardiac, vascular and thoracic surgery patients concluded that the maximum risks
of epidural haematoma were 1:1700, 1:1700 and 1:1400 respectively (Ruppen et al, 2006b
Level IV). However, this was a calculated risk only; there were actually no cases of epidural
haematoma reported in the studies (14 105 patients) used in this analysis.
In another analysis of a case series after cardiac surgery, the risk of epidural haematoma was
calculated at 1:12 000 (95% CI 1:2100 to 1:68 000) — comparable to an obstetric population.
It was described as being in the same risk range as receiving a wrong blood product (or the
yearly risk of having a fatal traffic accident in a Western country) (Bracco & Hemmerling, 2007
Level IV).
Early diagnosis and, if indicated, immediate decompression (less than 8 hours after the onset
of neurological signs) increases the likelihood of partial or good neurological recovery
(Horlocker et al, 2010 Level IV). CHAPTER 7
Epidural abscess
Serious neuraxial infections following epidural anaesthesia have previously been reported as
rare. However, prospective studies have found rates in the range of 0.015% to 0.05% (Kindler et
al, 1996 Level IV; Rygnestad et al, 1997 Level IV; Wang et al, 1999 Level IV). It is of note that in the
studies with these high incidences, patients had long durations of epidural catheterisation; the
mean duration in patients with an epidural space infection was 11 days; no infection occurred
in any patient whose catheter was in situ for less than 2 days and the majority of patients were
immunocompromised (Wang et al, 1999 Level IV).
Only 5.5% of 915 cases of epidural abscess published between 1954 and 1997 developed
following epidural anaesthesia and analgesia; 71% of all patients had back pain as the initial
presenting symptom and only 66% were febrile (Reihsaus et al, 2000 Level IV). The classic triad of
symptoms (back pain, fever and neurological changes) was present in only 13% of patients
with an epidural abscess (in a study unrelated to epidural catheterisation); diagnostic delays
occurred in 75% of these patients and such delays led to a significantly higher incidence of
residual motor weakness (Davis et al, 2004 Level IV).
Audit data from the study referred to above (Cameron et al, 2007 Level IV) showed that of the
8210 patients with epidural catheters over the period of 16 years, six developed epidural
abscesses. Only one of these required surgical decompression and they did not suffer any
long‐term neurological loss. The authors stress the importance of appropriate patient

Acute pain management: scientific evidence 187