Epidurals 'safer' than thought

“Epidurals and spinal anaesthetics are much safer than previously realised,”_ The Times_ reported today. It said the first nationwide census of the procedures has calculated the risks to be much lower that previously thought. The newspaper said that researchers had estimated that women having an epidural during labour only had a one-in-80,000 chance of suffering permanent harm, and that this was possibly as low as one in 300,000. It said that even high-risk patients, such as the frail and elderly, had between only one in 6,000 to one in 12,000 risk of permanent harm.

These figures come from an audit of a range of anaesthetic procedures, including epidurals. The thorough study collected reports from all NHS hospitals thought to be performing these procedures. It also monitored complications for a whole year, and referred to other sources of information to verify the findings. As such, these figures give a good estimate of the rate of complications of these procedures.

Where did the story come from?

Dr Tim M. Cook and colleagues from the Royal United Hospital in Bath carried out this research as part of the Royal College of Anaesthetists Third National Audit Project. The work was funded by the Royal College of Anaesthetists.

The study was published in the peer-reviewed British Journal of Anaesthesia . A complete report of the project has been published on the website of the Royal College of Anaesthetists, but this has not been reviewed here.

What kind of scientific study was this?

This was a national audit looking at the number of central neuraxial block (CNB) procedures performed in the UK annually, and the rate of major complications associated with the procedure. CNBs, which include epidurals, involve anaesthetising the lower half of the body by injecting anaesthetic into the space or fluid surrounding the spinal cord. This procedure is performed for pain relief during childbirth, and for other reasons. Although major complications such as paraplegia can result from CNBs, it is not clear how commonly these complications occur. Consequently, the Royal College of Anaesthetists set up an audit to determine the rate of these complications in the UK.

The researchers asked all the anaesthetic departments in NHS hospitals to take part between March and September 2006. Each department nominated a person to record all CNBs performed in their hospital over a two-week census period from around the end of September 2006. Failed attempts at CNBs were not recorded. The different types of CNB were classified as epidurals, spinals, combined epidural-spinals, and caudals. The reason for performing the CNB was also recorded: adult or paediatric perioperative (related to surgery), obstetric (childbirth), or chronic pain relief. Whether the procedure was performed by a non-anaesthetist was also recorded. The recorder also assessed their data as either “accurate”, a “close estimate”, or an “approximate estimate”.

The data from each department was added up and multiplied by 25 (this multiplication factor was based on the annual results of a large district general hospital). This calculation gave an estimate of the number of CNBs performed annually in the NHS.

The researchers used a similar system to identify all complications arising from CNBs over a one-year period, from September 2006 to August 2007 (with reporting encouraged until March 2008). They also encouraged reports of complications from clinicians of any specialty. Complications from failed attempts at CNB were recorded. Major complications included those with potential for serious harm to the patient, such as infections, haematoma, nerve damage or cardiovascular collapse. The reporters were also asked to record cases where an injection intended for one route of injection inadvertently ended up being injected by the wrong route (e.g. a drug intended for the epidural space ended up being injected intravenously), even if no harm occurred.

All reports of complications were reviewed by an expert panel, which decided on the likelihood that the complication was caused by the CNB (five categories ranging from “certain” to “no link”). The panel also assessed the severity of the complication, and the outcome of each case at six months or later. The panel used a standard method (the National Patient Safety Agency severity of outcome scale) to classify the severity of the initial injury and its outcome. They identified any cases of permanent injury, which was defined as symptoms lasting more than six months. They also identified any cases of paraplegia or death. As there was some subjectivity in deciding the cause and outcome of a complication, the panel classified cases either as a pessimistic/worst-case scenario view, or an optimistic/best-case scenario view.

The researchers cross-checked their figures against various national databases to see if their figures were correct. These included the National Reporting and Learning Service (NRLS), the NHS Litigation Authority (NHSLA), the Department of Health Hospital Episodes Statistics, the National Obstetric Anaesthesia Database, and the Medical Protection Society among others. In addition, the internet and medical journals were checked for reports of relevant cases, and the relevant individuals contacted for information as necessary.

What were the results of the study?

The researchers obtained reports from all of the hospitals that were asked to participate, with the majority of the hospitals (92%) rating their data as “accurate”. Using the results of their census, the researchers estimated that 707,455 CNB procedures are performed in the NHS annually. Just under half of these procedures (46%) were spinal procedures, 41% were epidurals, 6% were combined spinal-epidural, and 7% were caudals. The most common reason for the procedures were obstetric (45%), closely followed by reasons related to surgery (44%). Less common reasons for a CNB included treating chronic pain (6%), while 3% of procedures were in children, and about 2% were performed by non-anaesthetists.

Of the 108 possible complications reported to the panel, 84 were considered relevant for review, and 52 met inclusion criteria. After reviewing national databases such as the NRLS and NHSLA, as well as the medical literature and internet, the researchers identified one case of a wrong injection route that had not been reported. But this was the only case. None of the major complications were in children aged under 16 years, and most complications occurred in people aged over 50 years.

The subjective expert panel judged that, in a worst-case scenario (i.e. pessimistic), 30 of these complications could be described as permanent injuries. Alternatively, they judged that in a best-case scenario (i.e. optimistic), 14 could be described as permanent injuries. This meant that for every 100,000 CNBs there were 4.2 permanent injuries (pessimistic estimate), or 2.0 permanent injuries (optimistic estimate). In the pessimistic scenarios, for every 100,000 perioperative CNB procedures, there were eight permanent injuries, compared with 2.5 for chronic pain CNB procedures, 1.2 for obstetric CNB procedures, and zero for paediatric and non-anaesthetist procedures. During the year that was reviewed, there were (at worst) 13 deaths or paraplegias, equivalent to 1.8 events per 100,000 CNBs. Optimistically, there were five deaths or paraplegias, equivalent to 0.7 events per 100,000.

What interpretations did the researchers draw from these results?

The researchers concluded that “the data are reassuring and suggest that CNB has a low incidence of major complications, many of which resolve within six months”.

What does the NHS Knowledge Service make of this study?

This study thoroughly assessed the frequency of complications arising from CNB procedures in the NHS, and gives an indication that they may not be as common as once thought. There are a number of points to note:

• The accuracy of the estimates in this paper depend on the completeness of the reporting from each department, and also how reliably the cases were identified as being related to the CNB procedure both at the local and panel stages. There was a high response rate to the census, and external sources were checked in order to validate the data and identify unreported cases. This increases confidence in the results of the study.
• However, the authors note that their calculated incidence of complications should be seen as a minimum estimate because cases that were not reported, or were incorrectly excluded, would increase the rates.
• Using a two-week period to estimate CNB procedures for the whole year may have lead to some inaccuracy. However, the authors report that any error in this figure is likely to be small because all NHS hospitals provided data, and most of them rated their data as accurate rather than as an estimate.
• Although this information is representative of the NHS in the UK, it may not be representative of non-NHS institutions or other countries. In addition, as practices may change over time, these figures may not apply to other time periods.
• Although no major complications were found in children under the age of 16, this does not mean there is zero risk of major complications. Rather, CNB procedures were least common in this group (only 21,500 procedures) so rare risks may not have been picked up.

Overall, despite the potential problems with collecting accurate data, this study represents the best data on the rates of complications in CNBs. This should be reassuring to both anaesthetists and patients who may need these procedures.

Sir Muir Gray adds...

Anaesthetists have really got it sorted.