Article Text
PDF
Abstract
Some evidence suggests that psychogenic non-epileptic seizures (PNES) are associated with increased mortality. The authors obtained death certificate information in a cohort of 260 patients who presented with PNES between 1999 and 2004. The follow-up period averaged 7.92 years, during which 17 patients died, 12/17 were under the age of 75 years, giving a premature (<75 years) mortality rate of 0.58%, compared with a Scottish mortality rate for the 40–75 years age group of 0.41% per year. The main predictor of death was the patient age at presentation of PNES. There was no correlation with withdrawal of anticonvulsant medication, and death certificate information provided no evidence to suggest that the deaths were related to the seizure disorder. The population had a high mean deprivation rank, providing a possible explanation for a modest increase in premature death rate.
- Psychogenic non-epileptic seizures
- death
- mortality
- epilepsy
- surgery
- somatisation disorder
- head injury
Statistics from Altmetric.com
Introduction
Psychogenic non-epileptic seizures (PNES) or pseudoseizures are common: prevalence estimates vary from 2 to 33 per 100 000,1 with an incidence of up to 6.17/100 000/year.1 2 Patients with PNES make up approximately one quarter of the referrals to specialist epilepsy clinics.3
Some evidence suggests that PNES is associated with mortality. Iatrogenic death due to pseudostatus has been documented.4 A Scottish study, reporting the outcome of medically unexplained symptoms in 1030 patients,5 found that five of the eight patients, who had died during the study, had the initial diagnosis of PNES. During a trial of cognitive behavioural therapy in PNES, one death from suicide occurred in 66 patients over a 6-month period.6 While the number of patients in these studies is small, the results raise a concern that patients with a diagnosis of PNES may have a higher than expected mortality rate and that some of the increased mortality rates may be related to a missed diagnosis of epilepsy. While there is evidence that anticonvulsant drugs (AED) can be withdrawn safely in patients with PNES in relatively short term,7 there is no data from long term follow-up.
Methods
Between March 1999 and August 2004, 260 consecutive patients seen at our PNES clinic were included in an observational study.8 One hundred ninety-six patients (75.4%) were females. Mean age at first attendance was 37.8 years (range 13–87 years, SD 14.2) and mean age at onset of PNES was 30.8 years (range 6–70 years, SD 14.2).
The diagnosis was confirmed by video EEG (249 patients), or by ambulatory EEG (one patient) of events confirmed as typical by eyewitnesses. In the remaining 10 patients, events were either directly observed in the clinic, or eyewitness descriptions were typical of PNES, and AED successfully withdrawn with appropriate post-withdrawal monitoring.
The baseline variables were8: age at presentation, age at onset of PNES, gender, presence of epilepsy and learning disability, previous psychiatric referral and diagnoses, self harm, sexual abuse or other traumatic experiences, medically unexplained symptoms other than PNES, PNES frequency and type, employment status, social deprivation index determined by postcode,9 anticonvulsant and antidepressant drug therapy. In the course of a subsequent audit of long- term outcomes, we recorded deaths that had occurred in our original cohort of patients. We then acquired death certificate information from the Registrar General's office.
Statistical analysis
Statistical analysis was carried out using SPSS V.15.
The Mann–Whitney U test was used to compare continuous variables and the χ2 test was used for categorical variables. Simultaneous logistic regression models were used to evaluate the ability of independent variables to predict outcomes. Exploratory bivariate analysis was carried out for each variable. Independent variables correlating with outcome variables, at the 10% level or less (p>0.1), were considered for entry into the model. When screening for co-linearity identified two independent variables correlating at the 30% level (p<0.3) or less, the variable correlating less significantly with the dependent variable was eliminated. The remaining independent variables were entered into an initial model. Independent variables without significant predictive value at the 5% level (p<0.05) were then eliminated, and the final analysis was carried out.
Ethics
The baseline and 6–12-month follow-up data were acquired as part of a study that had ethical approval from the Southern General Hospital Research Ethics Committee.8 In Scotland, death certificate information is public.
Results
Up to 2010, 260 patients had a mean follow-up of 7.92 years (range 6.00–10.10 years). During the follow-up, 17/260 patients died, representing a mortality rate of 0.82% per year. Age at death ranged from 35 to 88 years (mean 73.0±14.3), and 12 patients were under the age of 75 years. This showed a premature (<75 years) mortality rate of 0.58% per year.
The certified primary causes of death were malignancy (7), pneumonia (5), cardiac (2), renal failure (1), hepatic failure (1) and drug overdose (1).
Secondary causes were cardiac disease (2), pneumonia (1), chronic obstructive airways disease (2), diabetes (1), liver failure (1), peripheral vascular disease (1), dementia (1) and alcohol dependence (1).
Patients who had died were older at presentation than those who had not died, at 57.2 (±14.3) years versus 36.0 (±12.9) years, and had a later onset of PNES, at 46.4 (+15.8) years versus 29.4 (±13.2) years (p<0.001 in both cases). There were no significant differences in any of the other variables collected. In particular, patients who had died were no more or less likely to have been withdrawn from AED (9/17 versus 129/260, p=0.790), or to have a final diagnosis of epilepsy plus PNES (0/17 versus 26/260, p=0.171). Our PNES population had a mean deprivation rank of 1879, just within the lowest 30% of the Scottish population.
Patient age at presentation and age at onset co-varied significantly, and age at onset was eliminated from the binary logistic regression model. This done, two variables significantly predicted death during the follow-up period: age at presentation was a positive predictor, and self-harm was negative. With each increase in age, at presentation of 1 year, the probability of death during the follow-up period increased by 1.111 (95% CI 1.059 to 1.165; p<0.0001). This equates to an increase in probability of 2.865 for every 10 year increase in age at presentation. Patients who had a baseline history of self harm were 14.618 times less likely to have died during the follow-up period (OR 0.196, 95% CI 0.054 to 0.725). However, this was only true if the age at presentation was in the same model, thus suggesting that the effect only operated when variance due to age was taken into account. Age at presentation did not significantly predict self-harm.
Discussion
The premature mortality we found in patients with a diagnosis of PNES was 0.58% per year, compared with mortality in Scotland for the 40–75 years age group of 0.41% per year.10 Our results therefore suggest that a diagnosis of PNES may be associated with a modest increase in mortality.
Serious physical illness is highly prevalent in a subset of PNES patients, with an onset of their attacks after the age of 55 years; anxiety, associated with these physical complaints may be implicated in the pathogenesis of PNES.11 Four of our patients who died were known to have significant health problems before diagnosis of PNES, neurological in two cases (cerebral palsy and posterior fossa tumour), pulmonary in the others. In these cases, it is possible that the added stress and distress associated with physical illness may have contributed to the development of PNES.
Anxiety and depression are associated with increased mortality,12 and are common in patients with PNES.7 We found no evidence of a predictive effect of psychiatric state in our population. The negative predictive effect of self-harm (which included suicide attempt in our dataset) was probably due to the known tendency for impulsivity to decline with age,13 and provided further evidence against a ‘psychiatric’ mortality.
A minority of patients with PNES (10%–15%) also have epilepsy.14 One obvious anxiety is that we had failed to detect underlying epilepsy in patients withdrawn from AED, and that some continued to have seizures and died from them. In most of our patients, the causes of death were plainly non-central nervous system and would be incompatible with this. In the case of patients who died of brain tumour, cerebral palsy and ‘pulmonary oedema’, we considered it conceivable that the deaths might be seizure related. The patient, who died of a primary brain tumour, was known to have it at the time of PNES diagnosis. It was situated in the posterior fossa, a site not normally associated with seizures. In all three patients, typical PNES had been recorded using video EEG, and in two patients, medication was withdrawn with our usual post withdrawal monitoring,6 with no emergence of epilepsy. The patient who died of pulmonary oedema had remained attack free since the diagnosis of PNES.
Finally, our patients came from a population that was deprived relative to the general Scottish population: some excess mortality might be expected secondary to socio-economic factors.
Our data suggests that any excess mortality in patients with PNES is modest, and provides no evidence of an effect of missed diagnosis of epilepsy, nor of withdrawal of AED.
References
Footnotes
Competing interests None.
Ethics approval The ethics approval was provided by Southern General Hospital Research Ethics Committee.
Provenance and peer review Not commissioned; externally peer reviewed.