Secondary Logo

Journal Logo

Case Report

A clinical diagnosis of inadvertent epidural administration of potassium chloride

Peduto, V. A.; Mezzetti, D.; Gori, F.

Author Information
European Journal of Anaesthesiology: June 1999 - Volume 16 - Issue 6 - p 410-412



The risk of the accidental introduction of substances other than the intended drug is always present when conduction anaesthesia is performed. However, the incident is usually discovered by chance or after symptoms prompt a careful review of the circumstances. We present a case of inadvertent epidural administration of potassium chloride (KCl) 15% in which the correct diagnosis was made on clinical signs.

Case report

A 65-year-old white male was referred to our pain therapy service for recurrent episodes of right lumbosciatica. An epidural injection of local anaesthetic with corticosteroid was planned. With the patient in a sitting position, a subcutaneous wheal was made at the L3-L4 level using mepivacaine 2%. A 18-gauge Tuhoy needle was inserted and the epidural space identified by the loss of resistance technique using 'saline'. After about 3 mL of the solution had been injected, the patient began to complain of dermatomal burning pain at the lumbar level, radiating to the legs. The administration was immediately terminated and the needle removed. A few minutes later, tachycardia (120 beats min−1), hypertension (190/110 mmHg), and diaphoresis appeared. The pain grew steadily worse and then diminished. Thirty minutes after the injection there was paralysis of the lower limbs with isolated clonic contractions, knee and ankle areflexia, complete anaesthesia up to T10 and urine incontinence. A neurologist was consulted, and an attempt was made to obtain a somatosensory evoked potentials (SSEP) tracing. During the recording, clonic contractions of the lower limbs rapidly increased in frequency and intensity. This hampered the recording of repeatable potentials, but suggested the basis for a differential diagnosis. The hypothesis that the clinical picture was caused by a neuroexcitatory substance was considered. The vials were recovered and the substitution of saline with potassium chloride 15% (2 mM mL−1) was noted. The solutions were contained in identical vials, even although they were marked with differently coloured labels, blue for saline and purple for KCl.

Midazolam (5 mg i.v.) was given to block clonic manifestations. Hydrocortisone (2 g) was administered to prevent possible medullary oedema. Ten minutes later the contractions disappeared and, in the next 4 h, progressive recovery of neurological function was recorded. A complete neurological examination at this time revealed no detectable anaesthesia and almost complete recovery of leg movements and reflexes. Two days later the patient was discharged without sequelae. After 7 days a SSEP of lower limbs showed normal latencies after stimulation of both the tibialis posterior and the peroneus communis nerves.


There are six descriptions (eight patients in total), to our knowledge, of the accidental introduction of KCl into the epidural space [1-6] in the international literature. The symptoms in the present case follow closely those already reported: pain, paresis, clonic contractions of affected limbs and signs of autonomic hyperactivity.

Examination of published data together with our report permits several observations. Pain may be present on injection if the KCl solution is used to identify the epidural space [6], otherwise it begins after a few minutes. Local anaesthetics introduced before KCl injection may delay the onset of symptoms up to resolution of their effect [4], but opioids do not [3,5].

The correlation between dose and duration of symptoms, is shown in Table 1. In effect, time to the beginning of resolution has a linear relation with the dose (R=0.946, P=0.0042, if approximate time reported by Lin [3] is considered as the mean of the two values). Doses less or equal to 1500 mg have been always followed by complete recovery, but paraplegia has been reported following a dose of 2250 mg. This amount may possibly represent the neurotoxic threshold.

Table 1
Table 1:
Relation between dose, concentration and effect in reported cases of epidural KCl administration

During a crisis, if the diagnosis is not obvious clinical interpretation must be undertaken. In the present case, the differential diagnosis addressed pain and motor/sensory block and included epidural/subdural haematoma, subarachnoid haemorrage and the injection of a neurolytic substance. These complications of epidural anaesthesia were excluded when clonic contractions of the legs developed. The absence of significant involvement of brain function made plausible their spinal origin.

Spinal seizures are an uncommon phenomenon. They have been linked with intrathecal injection of contrast media [7,8] and concentrated morphine [9]. Leg spasms have been reported in other reports of KCl epidural administration [3,4]. Animal models indicate that muscle relaxants injected into the cerebrospinal fluid may provoke seizures [10]. All these substances are readily available in the operating theatre and the anaesthesiologist should be aware of this diagnostic manifestation as an aid to diagnoses.

Somatosensory evoked potentials are undoubtedly useful in the evaluation of acute spinal injury and we used them to assess completeness of recovery. Such a test obtained during the acute stage would certainly have been of interest.

As a precaution, drugs which are used infrequently should not be kept in the operating room, but in a nearby location. The act of deliberately retrieving a drug may lessen the probability of error.

A recognized preventive strategy requires the adoption of an ampoule labelling standard that forces reading of essential information [11]. A fixed colour combination that enhances legibility (e.g. black on yellow), large type and a label arrangement that efficiently utilizes ampoule space may greatly affect legibility. Specifications regarding drug labelling already exist [12] and an effort should be made to apply them consistently. The vials utilized in this case were colour-coded, but typeface and legibility were poor and this may have contributed to the human error.


1 Pagani I, Carnevale L, Bonezzi C, Preseglio I. Descrizione di un caso clinico di errata somministrazione peridurale di potassio cloruro al 15%. Minerva Anestesiol 1984; 50: 407-410.
2 Shanker KB, Palkar NV, Nishkala R. Paraplegia following epidural potassium chloride. Anaesthesia 1985; 40: 45-47.
3 Lin D, Becker K, Shapiro HM. Neurologic changes following epidural injection of potassium chloride and diazepam: a case report with laboratory correlations. Anesthesiology 1986; 65 (210): 2.
4 Tessler MJ, White I, Naugler-Colville MA, Biehl DR. Inadvertent epidural administration of potassium chloride. A case report. Can J Anaesth 1988; 35: 631-633.
5 Liu K, Chia YY. Inadvertent epidural injection of potassium chloride. Report of two cases. Acta Anaesthesiol Scand 1995; 39: 1134-1137.
6 Vercauteren M, Saldien V. Epidural injection of potassium hydrochloride. Acta Anaesthesiol Scand 1996; 40: 505.
7 Maruyama K, Setoguchi Y, Maruyama J et al. Intrathecal injection of high-dose meglumine amidotrizoate with complete recovery. Intensive Care Med 1993; 19: 232-234.
8 Rivera E, Hardjasudarma M, Willis BK, Pippins DN. Inadvertent use of ionic contrast material in myelography: case report and management guidelines. Neurosurgery 1995; 36: 413-415.
9 Groudine SB, Cresanti-Daknis C, Lumb PD. Successful treatment of a massive intrathecal morphine overdose. Anesthesiology 1995; 82: 292-295.
10 Szenohradszky J, Trevor AJ, Bickler P et al. Central nervous system effects of intrathecal muscle relaxants in rats. Anesth Analg 1993; 76: 1304-1309.
11 Nunn DS, Baird WLM. Ampoule labelling. Anaesthesia 1996; 51: 1-2.
12 Standard specifications for labels for small-volume (100 mL or less) parenteral drug containers D4267. In: Annual Book of ASTM Standards. Philadelphia: American Society for Testing and Materials, 1989.

ANAESTHETIC TECHNIQUES, epidural; COMPLICATIONS, accident; IONS, potassium chloride; SPINAL CORD, seizures

© 1999 European Society of Anaesthesiology