Respiratory arrest with remifentanil patient-controlled analgesia – another case


  • Published with the written consent of the patient. No external funding and no competing interests declared. Previously posted at the Anaesthesia Correspondence website:

We would like to describe a case very similar to that of Bonner and McClymont involving a respiratory arrest within five minutes of commencing remifentanil patient-controlled analgesia (PCA) for labour analgesia [1].

A fit, primiparous woman requested epidural analgesia at 4–5 cm cervical dilatation following post-dates induction of labour and a Syntocinon® infusion. One litre of intravenous Hartmann’s solution was commenced and the epidural sited uneventfully. A test dose of 10 ml bupivacaine 0.1% with fentanyl 2 μ−1 was given without incident. Approximately 45 minutes later, the patient remained distressed and a second and third 10-ml dose of the same solution was required before analgesia was achieved. Approximately 90 minutes following insertion and despite a good sensory block bilaterally to T9, the patient described a sudden onset of left knee pain attributed to radicular impingement by the epidural catheter, which was promptly removed.

A trial of remifentanil PCA was suggested, to which the patient agreed. Two milligrams of remifentanil was diluted in a 50-ml syringe with saline 0.9%, and the giving set purged to avoid delivery of a remifentanil bolus, before connection to a functioning intravenous infusion incorporating unidirectional valves. A pre-programmed regimen was commenced (40 μ−1, 1-ml bolus (40 μg), 2-min lockout). Within five minutes of the patient’s successfully self-administering her first dose, however, she became unconscious and had a respiratory arrest, with cyanosis but no loss of cardiac output. She was immediately placed in the left lateral position with her airway supported manually, and given 100% oxygen. Spontaneous respiratory effort returned within 30–60 seconds, the patient becoming well perfused, shortly after which she regained consciousness. Remifentanil intoxication was assumed, and the PCA disconnected. Interrogation of the PCA machine showed that 160 μg had been delivered over four successful attempts.

Having considered other explanations, we suggest that the patient may have received an opioid overdose secondary to residual epidural fentanyl, and/or her intravenous infusion had temporarily obstructed causing bolus dosing of remifentanil, and/or her distressed hyperventilation had reduced her PaCO2 such that additional remifentanil had abolished any respiratory drive.

As a consequence, we have altered our remifentail PCA protocol using designated cannulae for PCA administration and stipulating that remifentanil should not be administered to any patient receiving epidural fentanyl within the preceding four hours. Like Bonner and McClymont, we have also deemed continuous oxygen saturation and designated one-to-one midwifery care to be mandatory, and suggest that although remifentanil PCA is a very effective method of labour analgesia, it requires very tight safety protocols to be in place.