Case

Decades ago, an 11-year-old girl with a history of previous bilateral clubfoot surgery as an infant was scheduled for unilateral corrective osteotomy at the foot level. After an inhalational induction a size 3 classic type laryngeal mask airway (LMA) was inserted and a proximal sciatic nerve block using the lateral approach was performed in the spontaneously breathing child. An advanced skilled resident was guided by the senior anaesthetist. The right foot was taped in a slight inwards rotation in order to elevate the greater trochanter, the position of which was then marked on the skin, and after extensive skin disinfection a 10 cm long isolated nerve stimulator needle was advanced below the greater trochanter in parallel to the surface of the table until the typical twitches appeared. The current was then turned down and 20 ml of bupivacaine 0.25% was injected at a site where, with 0.3 mA using an impulse width of 1 ms, twitches were no longer visible. Photos were taken during the procedure for future teaching purposes, including one showing the happy and proud resident. The right hip area was cleaned, the patient transferred from the induction room to theatre, and there the resident took further care of the patient.

Ten minutes later, he stuck his head out of the door of the theatre and said ‘everything is fine, but surgery is intended to be on the left side.’ Indeed, on the surgical list, a left-sided osteotomy was scheduled, as had been written on the anaesthetic chart at the time of the pre-anaesthetic visit. While the girl was still in theatre, the parents were seen on the ward by the anaesthetist and informed face to face about the mishap. At the end of the case a left-sided sciatic nerve block was performed, which provided long-lasting postoperative analgesia.

Discussion

This case shows that wrong-site blocks can occur, even in the context of scheduled routine surgery. A wrong-site block is a ‘never event’ and one of the most preventable medical errors. It is always a matter of negligence at some point. Nevertheless, the author was directly involved in this case and is aware of two other cases. In the literature an incidence in the range of 1–4 per 10 000 blocks has been reported (Hudson et al. 2015, Sites et al. 2014), predominantly involving femoral nerve blocks, and the topic has been the focus of review articles (Barrington et al. 2015). Naturally the risk of wrong-site blocks exists for pain blocks too (Cohen et al. 2010).

The most relevant risk factors are distraction, fatigue and unusual production pressure. In addition, as in this case, the blocked wrong extremity often shows some abnormalities, e.g. a cast, a scar or a malformation. It is of paramount importance not to perform the block on the apparently affected extremity, but to assure oneself that this existing pathology is indeed the target of surgery. In addition, every change in position, e.g. from supine to lateral or from lateral to prone, increases the risk of a right–left error.

For prevention it is important to check all documentation and to verify the site marking. The WHO surgical checklist has undoubtedly brought more stability to the perioperative process (Haynes et al. 2009). Recently a specific checklist for performing regional blocks has been proposed (Mulroy et al. 2014). The author uses a very simple spoken checklist before inducing general anaesthesia (cf. Case 2.1) and since this case a brief pause before inserting the block needle in order to check the side. Marking the surgical site was not yet standard practice in those days; a visible mark on the left foot would have undoubtedly prevented proceeding with a sciatic nerve block on the right side. In this case there was no hectic atmosphere and no distraction; the teaching senior anaesthetist saw the right foot with its abnormal appearance, and was so convinced that it must be the right side that the trainee didn’t question this.

The ‘stop before you block’ initiative recommends a pre-procedure pause to confirm the correct side for the regional anaesthetic block. The success of such an initiative relies on repeated education and necessitates a change in culture; especially in emergency cases and in out-of-theatre procedures, the site check often gets forgotten (Slocombe & Pattullo 2016). Unhappily, despite special efforts, the incidence of wrong-site blocks seems to have remained stable over the years (Pandit et al. 2017). To increase adherence to the ‘stop before you block’ checklist, a suggestion has been made to mark the block site and to use special stickers on the syringe (Chikkabbaiah et al. 2015). And recently the performance of a ‘mock block’, touching the skin with an empty syringe and so challenging the awake patient to confirm the side, has been suggested (Pandit et al. 2017).

Unfortunately, humans are only reliably protected against mistakes which they have already made themselves, and through which they have acquired their own hurtful experience. The author, as well as the former resident, now head of a department, still remember this story well, a little ashamed but with the passage of time also with a big smile. It is very likely that these two anaesthetists are well protected against performing wrong-site blocks in the future.

Summary and Recommendations

Wrong-site block is an ever-present threat, and every effort should be made to avoid it.

A visibly affected extremity may not be the target of surgery; one should never proceed without checking all the documentation and the marking on the skin.

A short break before inserting the needle and checking the side and the indication (‘stop before you block’) is strongly recommended.

Case

A long time ago, a 14-year-old boy with Duchenne muscular dystrophy, weighing 28 kg, was scheduled for bilateral tenotomy of the Achilles tendon and the tendon of the tibialis posterior muscle (needing an incision close to the medial malleolus). He had lost ambulation at the age of 8 and was on non-invasive BIPAP ventilation during the night since the age of 12. He had been seen in the outpatient clinic several weeks earlier, and a spinal anaesthetic was planned.

On the day of surgery, remifentanil 0.1 µg/kg/min was administered and he was positioned in the left lateral position for performing the spinal block; now an impressive scar was visible from the thoracic level down to the sacrum – obviously following scoliosis surgery and spondylodesis around a year ago. The anaesthetist requested to see the x-ray of the spine, but it was unavailable because this procedure had been carried out in another hospital. Several attempts with a 27G Whitacre needle using the paramedian approach at the L5/S1 level failed, as did an attempt to perform a caudal block. While the anaesthetist persevered with the regional block attempts, nitrous oxide, repeated boluses of propofol and finally sevoflurane were administered because the patient was obviously suffering pain. Finally a laryngeal mask airway (LMA) was placed and anaesthesia was maintained with sevoflurane up to 0.5 MAC and remifentanil. Bilateral ultrasound-guided distal sciatic nerve block was performed with 15 ml bupivacaine 0.25% on each side. Muscle twitches could not be elicited.

At the end of surgery, with the intention of speeding up the surgical list, the surgeon requested to wake up the patient, to send him to the ICU and to apply the reductive cast later in the afternoon, because he considered this be a non-painful procedure. However, redressing the foot was painful, and so the cast had to be applied later, again under general anaesthesia, now using a total intravenous technique.

Discussion

This case emphasizes several educational points:

First, the chosen regional technique should reliably cover the whole field of surgery. The tendon of the tibialis posterior muscle inserts at the navicular bone of the foot, the tendon is running behind the tibia, and the skin incision for tenotomy is made close to the medial malleolus. Until recently it was the predominant belief that the very variable saphenous nerve exclusively covered the skin, and that the deeper structures were all innervated by the sciatic nerve. However, recent evidence shows that the saphenous nerve also has branches to the periosteum of the tibia and the medial talocrural capsule (Eglitis et al. 2016). Therefore, to cover the whole foot, e.g. for hallux valgus surgery, fixation of an os tibiale externum, or as in this case a tibialis posterior tenotomy, both a sciatic nerve block and a saphenous nerve block, or more proximally a femoral nerve block, have to be applied (Fig. 5.2a). In this case, the non-blocked saphenous nerve most likely explains the pain felt during the trial to apply the cast with the patient awake.

Figure 5.2a A sciatic and a saphenous nerve block are both needed to guarantee a completely pain-free child after surgery involving the medial aspects of the ankle.

Second, spinal anaesthesia would have been an excellent choice in this high-risk patient with muscular dystrophy dependent on non-invasive ventilation during the night. Neuraxial anaesthesia can be successfully performed in patients with previous scoliosis surgery, although it is more demanding and has a higher failure rate (Ko & Leffert 2009). The responsible senior paediatric anaesthetist had successfully managed several such cases before. But in most of the reported cases in the literature scoliosis surgery did not involve the lower lumbar segments and the main difficulty was mastering the torsion of the spine. Choosing the L5/S1 interspace, the so-called Taylor’s approach, is usually successful, because at this level the distortion of the spine is still minimal. But in the case presented here the scar went down to the sacrum and information about the exact extent of the previous surgery was unobtainable. Nevertheless, the preoperative evaluation had been done some time in advance and it would have been part of good practice to request the spine x-rays from the other hospital, to study them, and to choose then the best interspace or to abandon the plan for spinal anaesthesia. Caudal anaesthesia can be done in adolescents too (Keplinger et al. 2016), and is always an option after spine surgery because this area usually remains untouched by surgery (Fig. 5.2b). As only the segments distal to L4 would need to be blocked, moderate volumes of local anaesthetics would have been sufficient. Unfortunately the anaesthetist failed to identify the sacral epidural space and abandoned the technique mainly because it caused discomfort to the patient; in addition, at that date, experience with ultrasound for neuraxial anaesthesia was still limited.

Figure 5.2b Caudal block can be successfully performed in most children with severe scoliosis. This photo shows another child; in the presented patient, unfortunately, this was not the case.

Third, patients with Duchenne muscular dystrophy present a high risk for anaesthesia, even when surgery is only minor. The disease is an X-linked recessively inherited dystrophinopathy with an incidence of 1:4000. It is the most common myopathy. Dystrophin is a large protein responsible for the integrity of the sarcolemma (Segura et al. 2013). It is completely absent in Duchenne muscular dystrophy and it has an abnormal structure in Becker muscular dystrophy, which has a slower progression. Patients with Duchenne muscular dystrophy not only have progressive muscular weakness, which is usually apparent by 2 or 3 years of age, but also cardiomyopathy. Cardiac involvement can also be present in female carriers of the afflicted gene. Life expectancy has dramatically changed over the last few decades: today, with home ventilation, survival into the forties is no longer unusual (Kieny et al. 2013). Therefore every anaesthetist now has the potential to be exposed to such a patient.

The defect in the dystrophin–glycoprotein complex causes chronic instability of the cell membrane which is further enhanced by the administration of succinylcholine and to a lesser extent by inhalational agents (Schieren et al. 2017). The administration of succinylcholine leads to rhabdomyolysis and massive, often lethal hyperkalaemia; this is a particularly serious problem in the early stages of the disease, when the diagnosis is often not yet known and the absolute amount of muscle is still significant. The potential of having patients with undiagnosed muscular dystrophy was the main driver in banning succinylcholine from elective paediatric anaesthesia. Succinylcholine is definitively contraindicated in patients with Duchenne disease. These children do not have an increased risk of malignant hyperthermia susceptibility compared to the normal population (Gurnaney et al. 2009), but occasional cases of rhabdomyolysis following sevoflurane, without the other typical features of malignant hyperthermia, have been reported (Simpson & Van 2013). There is some consensus that for an elective case total intravenous anaesthesia is the preferred choice; but, because difficulties with venous access are very common in this group of patients, the majority of paediatric anaesthetists would take the advantages of an inhalational induction (Brandom & Veyckemans 2013). It is noteworthy that the incidence of anaesthesia-induced rhabdomyolysis is much lower than the incidence of Duchenne disease, and in a large series of children presenting for muscle biopsies inhalational agents have been used uneventfully, although some of them later proved to have Duchenne muscular dystrophy (Flick et al. 2007).

In the presented case, the option to administer sevoflurane for a short time was not taken ‘voluntarily’, as a first choice, but because the situation evolved from slight sedation to general anaesthesia in a suffering patient, and sevoflurane was ready at hand. For the second anaesthetic, the classic approach of a total intravenous anaesthesia was chosen.

Summary and Recommendations

This case shows that for covering all deep structures of the foot not only a sciatic but also a saphenous nerve block is needed.

A complete preoperative workup including a clinical examination and looking at the radiological findings would have prevented the unsuccessful attempts to perform a spinal block.

Patients with Duchenne muscular dystrophy are at high risk for perioperative complications. Whereas succinylcholine must not be given, inhalational agents may be used if clearly beneficial, e.g. in case of difficult vascular access.

Case

Decades ago, a 10-year-old girl, weighing 33 kg, was scheduled for bat ear surgery. After an inhalational induction and the administration of atracurium the airway was secured with a size 6.5 preformed endotracheal tube. Anaesthesia was maintained with nitrous oxide and halothane. The area of the retroauricular skin incision was infiltrated with 8 ml bupivacaine 0.25% with adrenaline 7.5 µg/ml on each side in order to provide a bloodless field and postoperative analgesia. The perioperative course was uneventful; the child was extubated and transferred to the surgical ward.

About 1 hour later, on the postoperative visit, the child was found with a complete right-sided facial nerve palsy with the inability to close the right eye. These findings worried the nursing staff; surprisingly, the child and her parents accepted this as a normal phenomenon after bat ear surgery. Because the anaesthetic course was completely uneventful and no other neurologic deficit was present, a causal relation with the bupivacaine infiltration was thought to be the most likely explanation, and the decision was taken only to observe the patient; indeed, after 6 more hours the facial weakness had disappeared.

Discussion

This case of facial nerve palsy, which has been previously reported in part elsewhere (Jöhr & Sossai 1996), illustrates that any local infiltration extending to the mastoid carries the risk of a facial nerve block. This can be frightening for the parents and especially puts the child at risk of corneal erosions, as reflexes such as protective lid closure are impeded. It is often wise to forgo wound infiltration around the face, especially in the mastoid area or lateral to the eye, when the local anaesthetic is exclusively given for postoperative analgesia and the child has a general anaesthetic anyway. Facial nerve palsy has also been reported after infiltration at the neck, e.g. for carotid artery surgery (Hayek et al. 2003).

Accidentally blocking the vagus nerve leads to vocal cord paralysis, with transient hoarseness, or in case of bilateral block to respiratory insufficiency. This can happen high up in the neck with infiltration of the tonsillar bed for postoperative analgesia (Weksler et al. 2001) or during infiltration in the vicinity of the carotid sheath (Thermann et al. 2007). Spread to the phrenic nerve is a typical side effect of interscalene brachial plexus block, but it can be accidentally blocked by any undirected deep infiltration at the neck (Schiessler et al. 1989).

A Horner syndrome with ptosis, miosis, enophthalmos and ipsilateral hyperaemia occurs through a high thoracic sympathetic blockade, e.g. by epidural anaesthesia, or a direct spread to the stellate ganglion, e.g. following an interscalene plexus block. It causes per se no major problems except the unexpected appearance; but miosis or more important anisocoric pupils can create diagnostic uncertainty and trigger unnecessary investigations. Therefore, the anaesthetist in charge has the responsibility to inform parents and medical staff about the benign nature of this phenomenon. Because of the proximity of the anatomic structures and the loose connective tissue enhancing the spread of injected solutions, a Horner syndrome is quite common in paediatric patients. The author has seen a Horner syndrome after an ultrasound-guided infraclavicular block in a school-aged child.

A Horner syndrome is often seen in the context of epidural blockade, with a vast number of reports in the literature (Aronson et al. 2000). The main issue is that a bilateral Horner syndrome is almost impossible to diagnose, because anisocoric pupils, the predominant diagnostic criterion, are not present; therefore Horner syndrome is only realized in the context of an asymmetrical spread of the epidural block. There are reports of a late appearance of Horner syndrome during a continuous epidural infusion of local anaesthetics, often after many hours or even days (Courtman & Carr 2000). However, such reports probably do not reflect a sudden higher, and therefore worrying, spread of the epidural blockade on the affected side; rather, the blocked area during an epidural infusion naturally regresses over time and suddenly Horner syndrome disappears on one side because of an asymmetrical regression of the blockade. In the author’s practice, the presence of a Horner syndrome during well-functioning epidural analgesia was rarely a reason to reduce the infusion rate as long as motor impairment of the upper extremity was absent. The author has recently seen a Horner syndrome in an 8-month-old infant, weighing 6 kg, following a second caudal block with 1 ml/kg of ropivacaine 0.2% performed after urologic surgery with a delay of more than 3 hours after the first block (Fig. 5.3); and as usual, at first glance, unexplained anisocoric pupils at emergence frightened the anaesthetist.

Figure 5.3 A transient Horner syndrome in an infant after a second caudal block performed at the end of surgery.

In contrast to the transient nature of a Horner syndrome after local anaesthetics, stretching neural structures or a direct needle trauma can cause long-lasting symptoms. The author remembers, decades ago, a permanent Horner syndrome, or at least one that persisted over several months, after multiple trials to insert an internal jugular line. It has to be noted that moderate anisocoria is a physiological phenomenon; a pupil difference of 1 mm or more occurs naturally in 2.5% of children (Silbert et al. 2013). The author remembers several children who developed transient anisocoria during general anaesthesia without any signs of a neurologic pathology, perhaps in the context of an altered balance of the autonomic nervous system, and he has himself occasionally anisocoric pupils, especially when he is tired.

At the lower extremity, an unwanted epidural spread occasionally occurs with psoas compartment block (Schüpfer & Jöhr 2005). In addition, in the inguinal region the femoral nerve lies very superficial and can be blocked accidentally during ilioinguinal nerve block or wound infiltration (cf. Case 5.9).

Summary and Recommendations

This case illustrates the importance of ensuring that whenever local anaesthetic is injected at the head or the neck, no unwanted spread to the facial or vagal nerve occurs.

A Horner syndrome is a relatively common phenomenon, and it is usually recognized when only one side is affected. All caregivers have to be informed about the benign and transient nature of the phenomenon.

Case

Many years ago, a 2 3/12-year-old girl, weighing 12 kg, was brought to the children’s hospital after a fall from a window. She was whimpering, restless, showed irregular breathing and had an obviously distorted thigh. After the administration of 15 µg of fentanyl a skin wheal in the inguinal region just lateral to the pulsating artery was made and a 25G insulated needle was inserted. The nerve stimulator was set at 1.0 mA with an impulse width of 0.1 ms. No twitches were elicited, but after a clear give the tip of the needle was thought to be below the fascia iliaca, and after careful aspiration 12 ml of a 1:1 mixture of levobupivacaine 0.5% and prilocaine 1% was injected (2.5 mg/kg levobupivacaine + 5 mg/kg prilocaine). Soon afterwards, the child appeared to be relaxed and repelled the approaching nurse with a decisive ‘no’. The transfer to the CT scanner was prepared, but suddenly, 12 minutes after the injection, the child convulsed. The clonic convulsion subsided after administration of 20 mg of thiopental, and ventilation had to be briefly assisted with bag and mask. The decision was made to proceed with the spontaneously breathing but still unresponsive child to the CT scan for further diagnostics. The patient regained consciousness during transport. In addition to the femoral fracture, rib fractures and a small cerebral contusion area were found.

The femoral fracture was fixed by closed reduction and internal fixation under general endotracheal anaesthesia. Despite the positive-pressure ventilation and the presence of rib fractures, the team decided against inserting a chest tube because the thoracic region could be closely observed during the case. The child was extubated on table and the further course was uneventful.

Discussion

In this case the convulsions were likely to have been caused by local anaesthetic toxicity. The time delay of 12 minutes and the functioning analgesia from the femoral block indicate that rapid absorption and not a direct intravenous injection was responsible for the undesirably high plasma levels, although the doses used, with 2.5 mg/kg levobupivacaine, were in a high but still acceptable dose range. Prilocaine does not really add to toxicity, with the exception of generating methaemoglobin, which in this case was found to be at 3.6% after 90 minutes. Otherwise toxicity of mixtures of local anaesthetics is clearly additive; this has been shown for CNS toxicity as well as for cardiac toxicity. As no twitches could be elicited, the injection was probably made considerably lateral to the femoral nerve, similar to a fascia iliaca compartment block, below the fascia iliaca and probably always in part intramuscularly. When this block is performed with ropivacaine, a rapid absorption peaking as early as 10 minutes has been reported in children (Paut et al. 2004). As always, a diagnosis must be questioned: the coincidence of the small cerebral contusion and epilepsy would have been another possible explanation for the convulsions.

Local anaesthetic toxicity causes cerebral toxicity presenting with convulsions, and, more seriously, at higher doses cardiac toxicity. The treatment of cerebral toxicity consists in assuring oxygenation and interrupting convulsions by the administration of anticonvulsive compounds, e.g. benzodiazepines, thiopental or propofol. It has to be noted that a small dose of thiopental is usually sufficient to suppress convulsions; there is no need for a full induction dose, as for general anaesthesia followed by endotracheal intubation. In a case of severe cardiac toxicity, presenting with hypotension, conduction block and ventricular arrhythmias, the rules of advanced paediatric life support have to be followed; in particular, the administration of small doses of adrenaline is essential. In addition, lipid rescue is used, 1.5 ml/kg of lipid 20% followed by 15 ml/kg/h (Neal et al. 2012). The value of lipid rescue therapy has recently been questioned (Rosenberg 2016), but animal research as well as clinical cases underline its effectiveness (Weinberg 2017). Although the mechanism of action is still unknown, lipid rescue seems to play a major role in the treatment of the toxicity of local anaesthetics and other compounds (Presley & Chyka 2013).

Convulsions following the administration of local anaesthetics do occur; although in most cases CNS toxicity may be suppressed by the concomitantly administered general anaesthesia. In the famous first French ADARPEF study, including 24 409 regional blocks, two children convulsed (Giaufré et al. 1996). In the second ADARPEF study, including 31 132 regional blocks, 15 children showed signs of cardiac toxicity and one child convulsed in the group of 1262 pure regional blocks (Ecoffey et al. 2010). During his professional career the author was responsible for three cases in children, including the one described here. A school-aged girl convulsed during the injection of 1 ml/kg of 0.5% chloroprocaine for a Bier’s block augmenting a partially failed distal sciatic nerve block. Obviously the tourniquet on the lower leg was not sufficiently tight. And ages ago, the author performed bilateral axillary plexus blocks for the reduction of both fractured forearms in an adolescent with a severe midface trauma. Plastic cannulas had been placed blindly into the perivascular sheath and 0.5% bupivacaine was given incrementally; because the blocks did not develop satisfactorily this ended up as 80 ml. Then the boy convulsed. This case occurred in the very early years of the use of bupivacaine, when doses up to 6–8 mg/kg were thought to be still safe.

A femoral fracture is a typical injury for a preschool child; the incidence peaks at the age of 2, when three times more boys than girls are involved (Bridgman & Wilson 2004). Paediatric femoral fracture is completely different to the adult type. A minimal trauma, e.g. a simple fall on the floor, is sufficient and blood loss is negligible (Lynch et al. 1996). Therefore providing analgesia is the first thing to do and fluid administration is usually not needed. On the scene, nasal fentanyl, e.g. 1.5 µg/kg, is a good option followed as soon as possible by a femoral nerve block. Femoral nerve block is the only major conduction block which can be indicated outside the hospital in emergency medicine (Ronchi et al. 1989). Over more than two decades it had been the standard practice in the children’s hospital in Lucerne that every time a child with a suspected femoral fracture arrived, the anaesthetist was called to perform a femoral nerve block prior to removing the clothes and taking the x-ray. Unfortunately, because of new organizational structures, the enthusiasm to call a specialist has diminished.