Secondary Logo

Journal Logo


Inadvertent administration of hypertonic saline during postoperative thoracic epidural patient-controlled epidural analgesia

A case report

Bicket, Mark C.; Lester, Laeben C.; Kroll, Caleb E.; Brookman, Jason C.

Author Information
European Journal of Anaesthesiology: August 2016 - Volume 33 - Issue 8 - p 601-602
doi: 10.1097/EJA.0000000000000440


Among various pharmacological options for the relief of pain, hypertonic saline has been administered via the neuraxial route for analgesic benefit. Concentrations of epidural hypertonic saline for pain relief have ranged from 7 to 30%,1,2 with all applications focusing on chronic pain associated with malignancy or low back pain.1–3

The case report is unique in presenting the consequences of inadvertent administration of buffered 3% saline via the thoracic epidural route for analgesia following surgery. Outcomes associated with administration of buffered hypertonic saline, analgesia associated with this low concentration of hypertonic saline and epidural hypertonic saline use in the postoperative setting have not been described in previous reports. The patient consented to this case report.

A 63-year-old woman underwent endoscopy and laparoscopic Nissen fundoplication for refractory gastro-oesophageal reflux. Operative anaesthesia included general anaesthesia (isoflurane without nitrous oxide) and insertion of a thoracic epidural catheter at T6-T7. An infusion of 0.1% bupivacaine with fentanyl 10 μg ml−1 at a rate of 6 ml h−1 provided supplemental analgesia during the operation. Intraoperatively, intravenous fentanyl totalling 250 μg (given during the first 1.5 h of the 5-h procedure) was given. No other analgesics were used intraoperatively. Postoperative pain control with patient-controlled epidural analgesia (PCEA) was anticipated. Medication orders included 0.125% plain bupivacaine (4 ml h−1 continuous infusion, demand dose 2 ml, 10-min lockout interval, maximum six doses per hour). Following uneventful completion of surgery, the patient recovered in the post-anaesthesia care unit. Approximately 3 h after surgery, 18.5 ml of fluid had been infused via PCEA. No other analgesics had been given postoperatively. At this time, inadvertent administration of buffered 3% hypertonic saline via the epidural catheter was discovered.

Evaluation revealed an overall numerical rating scale for pain of 2/10. The patient was immediately informed of the error, and commented that she had never been ‘so comfortable’ after any of her previous operations. She denied neurological symptoms and back pain greater than her baseline chronic low back pain. Physical examination revealed no concerning neurological signs. Thirty minutes after stopping the saline infusion, numerical rating scale pain score increased to 7/10. Flushing the epidural catheter with 5 ml of 0.9% saline was followed by a 5 ml bolus dose of 1% lignocaine. The PCEA infusion was restarted at the above settings with a new bag of 0.125% bupivacaine with fentanyl, resulting in resumption of excellent pain control. Follow-up continued until her hospital discharge. The remainder of her hospital stay remained uneventful and she had no neurological complications at discharge.

A safety review revealed appropriate placement of the computerised order by the anaesthetist and correct receipt by the pharmacist who was responsible for preparation of epidural analgesic solution. The bupivacaine label was then printed and inadvertently placed in the pharmacy on an appropriately labelled 1000 ml bag of sodium acetate-buffered 3% hypertonic saline. The PCEA bag of this patient displayed labels for both bupivacaine and hypertonic saline. Chemical analysis by the pharmacy confirmed that the solution contained only buffered 3% hypertonic saline with 50% sodium chloride and 50% sodium acetate, with neither bupivacaine nor fentanyl. The hospital safety committee reviewed the error, and re-examined the medication labelling process.

Although epidural hypertonic saline has been used to treat chronic pain from epidural adhesions,1 few reports describe its use for acute postoperative pain (supplemental digital content, The safety of epidural hypertonic saline injections primarily stems from epidural lysis of adhesions, in which a solution containing hypertonic saline is injected to mechanically disrupt epidural scar tissue, and which is considered a well tolerated chronic pain procedure.1 Nerve irritation is not un-common following these injections, but typically resolves after several minutes to hours.1–3 Two non-lysis of adhesions hypertonic saline studies note few other complications.2,3

No immediate adverse events from hypertonic saline administration were noted in our patient. One explanation for lack of side-effects includes prior administration of local anaesthetic via the epidural route, which may have mitigated the possibility of pain or sympathetic stimulation when the hypertonic infusion was started. Past clinicians have suggested provision of local anaesthesia in the minutes prior to injecting hypertonic saline2,3 to avoid side-effects of hypertonic saline.

Another explanation for lack of side-effects may involve characteristics related to the delivery or composition of the solution. First, this patient's epidural space was probably subject to lower pressures during hypertonic saline administration because pressures associated with a low-volume infusion differ from direct boluses by hand via either the inter-laminar or transforaminal approach.3 Second, with a relatively lower concentration of saline compared with many existing studies, 3% hypertonic solution may not result in as significant perturbations as concentrations equal to or in excess of 10%. Finally, the buffered property of hypertonic saline that the patient received contrasts with non-buffered solutions of prior reports. Composed of 50% sodium chloride and 50% sodium acetate, the buffered mixture lowers the total amount and concentration of chloride in solution compared with non-buffered solutions.

Explanations for the analgesic effect of hypertonic saline have evolved over time, but focus on either osmolality or chloride ion concentrations. Initial hypotheses attributed interruption in Aδ and C fibre transmission to high osmolality,4 and studies of hyperosmolar solutions on nerve conduction have demonstrated neuromodulatory effects in animals.5,6 An alternate hypothesis centres on the high extracellular concentration of chloride.5 More recent investigations suggest that alterations of intra-cellular chloride ions may cause changes in pain signalling.7 However, our patient demonstrated a significant decrease in pain during the continuous infusion of buffered hypertonic solution, which has a lower chloride concentration. Based on her experience, significant analgesia may result from lower concentrations of chloride or lower osmolality than previously thought. Investigation of the safety of lower chloride concentrations in buffered hypertonic saline via the epidural route using animal models may justify additional examination of analgesia from hypertonic saline.

Acknowledgements relating to this article

Assistance with the letter: The authors thank Junichi Naganuma, MD, MBA, MHA, MPH for his assistance with this correspondence.

Financial support and sponsorship: none.

Conflicts of interest: none.


1. Jamison DE, Hsu E, Cohen SP. Epidural adhesiolysis: an evidence-based review. J Neurosurg Sci 2014; 58:65–76.
2. Kanayama T, Nakagawa H, Suzuki F. Epidural infusion of hypertonic saline solution for the relief of pain. Masui 1969; 18:1463–1468.
3. Koh WU, Choi SS, Park SY, et al. Transforaminal hypertonic saline for the treatment of lumbar lateral canal stenosis: a double-blinded, randomised, active-control trial. Pain Physician 2013; 16:197–211.
4. Shapiro H. Osmotic properties of frog nerve. Comp Biochem Physiol 1966; 19:225–239.
5. King JS, Jewett DL, Sundberg HR. Differential blockade of cat dorsal root C fibers by various chloride solutions. J Neurosurg 1972; 36:569–583.
6. Matsuka Y, Spigelman I. Hyperosmolar solutions selectively block action potentials in rat myelinated sensory fibers: implications for diabetic neuropathy. J Neurophysiol 2004; 91:48–56.
7. Price TJ, Cervero F, Gold MS, et al. Chloride regulation in the pain pathway. Brain Res Rev 2009; 60:149–170.
© 2016 European Society of Anaesthesiology