Postoperative pseudostatusNot everything that shakes is epilepsy

Authors


M. Reuber , Neurology

Abstract

Postoperative epileptic seizures are recognised but rare. Psychogenic seizures and pseudostatus epilepticus are relatively common, particularly in the peri-operative period. Our series of five cases of postoperative pseudostatus epilepticus demonstrates that the failure to recognise the psychogenic nature of this condition may cause anaesthetists to give inappropriate and potentially harmful treatment. Psychogenic ‘status’ is easy to diagnose once it has been considered. Convulsive episodes lasting longer than 90 s, closed eyes during a ‘tonic–clonic’ attack, retained pupillary response and resistance to eye opening are useful signs. Often there is a history of multiple admissions with ‘status epilepticus’ and of previous postoperative ‘status’.

Psychogenic seizures are common and often complicated by recurrent episodes of pseudostatus epilepticus [1]. Psychogenic seizures resemble epileptic attacks but are not associated with abnormal electrical discharges in the brain. It is estimated that up to 28% of patients referred to epilepsy centres with intractable seizures have nonepileptic attacks [2]. It is a misconception that psychogenic seizures often occur in patients with epilepsy. A study in which patients with presumably mixed seizure disorders were thoroughly investigated showed that only 10% had evidence of both disorders [3].

Psychogenic seizures are serious. Although patients with nonepileptic attack disorders consume considerable health resources their psychosocial needs are often not met. In one series, 67% of pseudostatus cases had made suicide attempts [4]. Furthermore, the treatment of psychogenic attacks with anticonvulsants is associated with a considerable morbidity and (probably) mortality. During one study which followed 13 pseudostatus patients over 4 years there were eight episodes of respiratory arrest caused by intravenous anticonvulsants [1].

Seizure-like movements frequently occur in association with general anaesthesia. There is experimental evidence that many anaesthetic drugs can lower the neuronal ‘seizure threshold’ [5, 6]. Despite this, postoperative epileptic seizures are extremely rare [7, 8]. Here we present a series of five cases of postoperative pseudostatus epilepticus. The fact that they were collected over a 6-month period from three hospitals in the Leeds area suggests that psychogenic seizures may be a comparatively common anaesthetic problem. Two cases are presented in detail, while all five are summarised in Table 1.

Table 1.  Pseudostatus epilepticus: summary of five cases Thumbnail image of

Case histories

Case 1

A 35-year-old former nurse was admitted to hospital for the excision of a labial sebaceous cyst. Her medication included phenytoin and vigabatrin. A past medical history of ‘known epilepsy’ with daily seizures was noted in her admission clerking. In the recovery room, the patient was thought to have developed status epilepticus with tonic–clonic seizures. She was treated sequentially with intravenous thiopental (30 mg), diazepam (10 mg), propofol (3 × 60 mg) and phenobarbitone (180 mg). After 30 min, there was still no response and a neurology registrar was asked for advice. The patient was in pseudostatus epilepticus. There were rolling head and markedly asynchronous limb movements; the patient resisted eye opening; she was not cyanosed. The abnormal movements settled quickly with reassurance and the patient ‘woke up’. Attacks of unresponsiveness and abnormal movements became gradually less frequent during the patient's postoperative 4-day observation on a ward.

The patient first developed episodes of impairment of consciousness when she was a 19-year-old nursing student. She was admitted to hospital several times in a limp, unconscious state lasting hours to days. Routine blood tests, EEG, head CT and lumbar puncture were normal. Eight months later, while working in an Accident and Emergency department, the patient developed attacks of nausea and a smell of ether followed by loss of consciousness and generalised shaking. Prolonged attacks precipitated numerous admissions to hospital. During one admission, the administration of diazepam 40 mg intravenously led to a respiratory arrest. The patient's general practitioner had already noticed that her seizures tended to get worse with the rectal administration of diazepam. Many of the observed attacks were not thought to be typical of epileptic seizures but, despite this, combination anticonvulsant therapy was given in high doses. Anticonvulsants were continued, despite two normal ictal EEG recordings and numerous normal postictal prolactin estimations because it was thought impossible to exclude the co-existence of epileptic attacks. Aged 24 years, the patient had a seizure during a tooth extraction under a local anaesthetic. A few days after this she had a series of seizures before a further extraction could be undertaken. One year later, the patient had a bilateral carpal tunnel decompression for relief of suspected median nerve compression, despite normal neurophysiology. There was an episode of postoperative ‘status epilepticus’ with seven ‘generalised seizures’ over 40 min treated by the attending anaesthetist with oxygen alone (no desaturation was noted). Three years later, the patient developed postoperative ‘status epilepticus’ after the excision of breast lumps under general anaesthetic. The lumps, which had been observed for several months, were found to be haematomas surrounded by inflammatory tissue and fat necrosis and recurrent breast trauma was considered to be the most likely cause. The patient's further past medical history included a seronegative arthralgic and myalgic disorder with reported (but not observed) episodes of arthritis. This condition has been treated with sulphasalazine and hydroxychloroquine. No radiographic evidence of erosions was detectable after 10 years of treatment.

Case 2

A 22-year-old former auxiliary nurse and lifeguard was admitted to hospital for the excision of a Bartholin's abscess. In the presence of a friend, she told the surgeons and anaesthetists who interviewed her before her operation that she had been investigated for epilepsy in the past but that a diagnosis of a nonepileptic seizure disorder had been made. She informed them that her seizures should not be treated with diazepam because this had made them worse in the past. The nurses gave her a red armband saying ‘No Diazepam!’. After removal of the tracheal tube, the patient felt light-headed and in pain. Shortly after she had arrived back on the ward, she developed apparent status epilepticus. When the fitting did not stop after diazepam (3 × 10 mg given intravenously), treatment with thiopental, atracurium besylate and propofol was commenced and the patient was intubated. She was transferred to an ICU 20 miles away because there were no intensive care beds available locally. Prior to her transfer, she was catheterised, a central venous and an arterial line were inserted. The patient was extubated the following day after her previous neurological records had been studied. Before she discharged herself the following day, three of her typical seizures were observed. She groaned and held her fists clenched against her chest while her legs shook. She was not cyanosed and held her eyes firmly closed; there was no postictal confusion.

This patient's attacks started at the age of 14. She was noticed to have hyperventilation attacks and apparently manipulative ‘faints’ on a paediatric ward to which she had been admitted for the investigation of headaches and light-headedness. During the same admission she had an attack of double vision which was thought to be hysterical. An EEG and head computed tomography (CT) were normal and she was diagnosed as having a hyperventilation syndrome. She received in-patient treatment in an adolescent psychiatry unit. At the age of 16, she had her first convulsive episode while at college. She was unresponsive, had clenched fists and shaking legs on arrival at hospital after 30 min. Diazepam (10 mg given intravenously) had no effect and a neurology registrar diagnosed pseudostatus epilepticus. Over the following 8 months, the patient had three to four seizures per day and there were at least five further admissions with pseudostatus. A prolactin and an EEG within 10 min of an episode of ‘status’ were normal. On one occasion the fits started in the family practitioner's surgery. The patient was briefly treated with anticonvulsants without beneficial effect. When aged 17 the patient was admitted to an epilepsy unit for a 2-month period of observation and treatment. An ictal EEG and postictal prolactin levels were normal. The diagnosis of psychogenic seizures was confirmed and the patient received psychological treatment. When an attempt was made to discharge her to her parents' home she took a deliberate paracetamol overdose. She disclosed a history of sexual abuse by her father, and later by her mother, to her female social worker. One year later the seizures continued at times of stress and in addition the patient exhibited self-destructive behaviour. Aged 20 she was started on carbamazepine treatment after she had developed postoperative seizures following a tooth extraction at another hospital. The drug was stopped after a deliberate overdose with it. Over the next year she could be fit-free for many months but would then have a period of daily fits. When reviewed again by a neurologist she revealed a history of prostitution from the age of 12 as well as amphetamine addiction. The patient has an additional past medical history of reported Still's disease (without residual symptoms or signs) and asthma.

Discussion

In all five cases a diagnosis of ‘known epilepsy’ was accepted too readily. Seizure observation of the postoperative attacks should have suggested a psychogenic cause. Several patients demonstrated side-to-side movements of the head and out-of-phase jerks of the limbs [9, 10]. The periods of generalised motor activity lasted much longer than in epileptic seizures in which they rarely exceed 90 s [9]. None of the patients showed cyanosis. All had closed eyelids during apparently generalised tonic–clonic seizures [11]. Several patients also showed resistance to eye opening, which is an even more conclusive sign [12]. When the pupillary response could be assessed, it was normal (as in around 90% of psychogenic seizures; 13). There was surprisingly swift reorientation after ‘generalised’ convulsions.

A response to irritant or noxious stimuli may sometimes be elicited in psychogenic attacks but is not present in generalised epileptic seizures. The release of the patient's arm over the head is often purposefully modified in psychogenic attacks. Importantly, urinary or faecal incontinence and serious injuries (including tongue biting) may occur in both types of seizure [14].

Several aspects in the patients' history could have set warning bells ringing. All five patients had been admitted to hospital with episodes treated as status epilepticus on more than one occasion. Recurrent status epilepticus is now very uncommon in the absence of a history of alcohol abuse or the type of neurological disease associated with abnormal imaging or EEG. Three patients had previous episodes of postoperative ‘status’. The semiology of the ‘fits’ had changed frequently. Seizures had become more violent with time. Anticonvulsants had made attacks worse. In common with many patients with psychogenic seizures, cases 2, 3 and 5 had an ‘unrelated’ psychiatric history [1] and a history of reported sexual abuse [15]. Patients 2, 4 and 5 had attempted suicide.

The previous investigations of all of the patients included ictal EEG recordings, which showed only movement artefacts during typical, ‘generalised’ seizures. This strongly suggests a nonepileptic seizure syndrome. Interictal EEG recordings may have been nonspecifically abnormal, a finding recorded in around 30% of psychogenic seizure patients [1, 15]. Postictal prolactin levels were normal. Cerebral imaging was normal despite the history of bad epilepsy.

Our cases show that a pre-operative neurological review may well be indicated if a patient gives a history of previous postoperative ‘status epilepticus’ or of several episodes of ‘status’ leading to emergency admission. If the history does not raise any diagnostic doubt, a combination of the easily obtainable observational and examination findings described here should produce a correct diagnosis of generalised epileptic or tonic–clonic-like psychogenic seizures in more than 90% of cases [9]. Both types of seizures are treatable if diagnosed correctly. It is usually not difficult to diagnose pseudostatus epilepticus once it has been considered. Our case series demonstrates that anaesthetists must be circumspect if they want to avoid giving inappropriate and potentially harmful treatments in the peri-operative period.

Ancillary