Opioid use in the bleeding obstetric patient: a clarification regarding NAP7

Abstract

Margiotta and Plaat offer an argued rationale on how general anaesthesia should be conducted for the obstetric patient who has bled extensively, suggesting that high-dose opioids may be in regular use and arguing that lower doses should be used [1]. Regarding NAP7, which the authors refer to, we offer some clarification.

In NAP7, of the 28 cases of obstetric peri-operative cardiac arrest [2], six involved both a general anaesthetic and obstetric haemorrhage. In two cases the haemorrhage occurred at some time after, and distant to, induction of general anaesthesia so these are not germane to the discussion. Of the four relevant cases, at induction, none received fentanyl: two received alfentanil and two no opioid. None received any other opioids before cardiac arrest. All four patients did receive vasopressors before cardiac arrest.

Regarding anaesthesia as a cause of cardiac arrest, as we explained in a recent letter [3] this is not equivalent to being an indication of iatrogenic harm. As an example, in the case of anaphylaxis, the causal factors are the patient (due to their allergy) and anaesthesia (as anaphylaxis would not occur without drug administration). It is no surprise that in obstetric anaesthesia, dealing with generally young and healthy patients, the patient is an infrequent cause of cardiac arrest. Further, as many obstetric anaesthetic interventions (e.g. neuraxial analgesia) involve no surgery at all, it is inevitable that in peri-operative obstetric cardiac arrest, anaesthesia will be a relatively more prominent cause than surgery. Thus, the observation that 68% of NAP7 obstetric cases have anaesthesia as a key cause is, in large part, a consequence of the nature of the service, rather than any indication of poor quality anaesthesia care. We separately assessed the quality of care (before, during, after cardiac arrest, and overall): obstetric anaesthesia had among the lowest ratings of poor care of all specialities in NAP7. Specifically considering the four cases cited above, none were judged to involve poor quality care by the assessing panel.

We note that the DREAMY study, the most recent large-scale service evaluation of UK obstetric general anaesthesia, reported that 57% of surgical obstetric cases received no opioids at induction, including 64% of caesarean sections [4]. It is likely that few obstetric cases are performed as a ‘cardiac anaesthetic’, particularly as NAP7 identified that most emergency obstetric anaesthesia is delivered by resident doctors.

As there were only four cases of obstetric haemorrhage and cardiac arrest in NAP7, the project can add little robust data to the debate. Certainly, although we did not collect drug doses, there was no evidence of widespread excessive opioid use at induction or of a classic ‘cardiac style’ induction. The NAP7 report does discuss the potential for induction drugs to cause haemodynamic compromise in a hypovolaemic obstetric patient and whether alternatives, such as ketamine, may be preferable [5]. Similarly, too high doses of induction drugs, such as thiopental and propofol, administered to patients in shock, were identified as a contributing factor to poor outcomes in the 2014 MBRRACE Report [6].

No doubt the discussion regarding general anaesthesia and the bleeding obstetric patient will continue. We hope this letter clarifies the relatively sparse data available from NAP7.