A reply

Thank you for giving us the opportunity to reply to the comments made by Doctors Mallik and Bhaskaran and Doctors Alston, Sinclair and Scott. Doctors Mallik and Bhaskaran ask about our informed consent. We gain informed consent from all patients in this hospital who undergo any procedure. This involves a description of the type of anaesthetic and surgery they will be receiving. It is the hospital's standard practice that all studies are presented to the Ethics Committee to gain approval before data from patients' medical notes are examined and analysed statistically and this was done for our retrospective study. Drs Alston, Sinclair and Scott worry about bias in the study in favour of the epidural technique, yet our demographic data clearly showed no differences between the two groups and our extubation criteria were well defined. We have only presented objective data. Our ‘natural enthusiasm’ was tempered by the deliberate omission of such subjective data as the degree of pain and its relief. This did not therefore appear in the paper and we are unsure as to why it is mentioned. As Drs Alston, Sinclair and Scott state, the risk of epidural haematoma in patients who receive epidural analgesia for coronary artery bypass graft surgery is unknown, but collective data have now been published on nearly 2000 patients without a documented haematoma. In addition, spinal and epidural haematoma can occur spontaneously even in patients who have never received neuraxial blockade or heparinisation [12345]. The well-documented risk of death (4%) and major complications such as postoperative myocardial infarction and cerebrovascular accident (4%) following this type of surgery is high enough that it requires to be fully explained to patients before they sign a consent form. We believe that the risk of epidural haematoma is much lower than 1% and does not require specific informed consent. Thus, put into context, if, for the sake of argument, the risk of haematoma is 1 in 1500 then for each haematoma that occurred, 60 patients would die and 60 would have a stroke as a direct consequence of the surgery. If it is possible to decrease the risk of common postoperative complications, with only a minimal added risk of epidural haematoma, then we feel the technique should be investigated further. Doctors Mallik and Bhaskaran state that they feel the catheter should be sited 12–24 h before the onset of surgery. Whilst we accept that some institutions do site their catheters earlier than we do, similarly there are other centres that site their catheters immediately prior to surgery [678]. We do not accept that the risk of bloody tap in the hands of experienced operators is in the order of 1–11%. Certainly in our experience of over 300 thoracic epidurals for cardiac surgery to date the incidence of bloody tap through the needle is 1 in 300. Our protocol states that in this situation the operation, and therefore heparinisation, is deferred as agreed in other studies [91011]. In the presence of normal clinical history and coagulation studies, we believe that any bleeding that may have occurred on insertion of the catheter, whether recognised or not, will have ceased prior to heparinisation [10, 12]. There has not yet been a large prospective study performed to look specifically at the differences between the two techniques. When examining relatively rare complications the power of any study involving small numbers is unlikely to be statistically useful. In addition, the abstract by Fillinger et al. quoted by Drs Mallick and Bhaskaran as evidence against an epidural technique is very misleading since it did not measure outcome data despite the title! Whilst we are familiar with the papers by Thorelius, Stenseth and Moore, we were not able to collect these data in our epidural patients, as it was a retrospective study. All other studies suggest that there may be benefits to be gained for patients by using this technique and currently we are performing a large prospective study to address this lack of information. We are also assessing some of these missing parameters. The concern over epidural haematoma has arisen in the literature because epidural techniques were developed initially in the obstetric population. In this situation where pain relief is the major advantage of the technique and mother and child have a combined life expectancy in the region of 120 years, a conservative approach to avoid even one case of epidural haematoma is only right and is completely endorsed and reinforced by ourselves. Major surgery, however, particularly in the elderly or high-risk population, carries significant morbidity and mortality much of which may in fact be diminished by regional sympathetic blockade [13] and cortisol inhibition, in contrast to the thoughts of Doctors Mallik and Bhaskaran. A review of the literature has, therefore, led us to believe that postoperative morbidity may be reduced by careful use of regional anaesthetic techniques. We do not wish to be seen to be complacent and every possible effort is made to minimise the progression of an epidural haematoma in our patients, should one develop. This can only be achieved by both the full support of one's surgical and nursing colleagues and careful and meticulous monitoring. We have strict protocols for neurological assessment and until further data are collected we would not support the practice without all of these criteria being fulfilled. That having been said, those of our anaesthetic colleagues who express such concern about so rare an event fail to appreciate that conventional analgesia with opioids carries a significant risk also, especially in the high-risk and elderly populations. Not one outcome study in over 100 in the published literature has demonstrated that narcotic analgesia is safer than, or associated with less morbidity than, regional anaesthesia [13]. Rather than being condemned as ‘regional enthusiasts’, we are keen to produce objective evidence for or against the use of either technique for this form of surgery. We believe we have tackled this controversy appropriately, namely an initial thorough literature search followed by a pilot group of patients and a subsequent large prospective study.