Epilepsy is a leading cause of disability and economic burden. Most of the evidence pertaining to the epidemiology of epilepsy is derived from hospital- or cross-sectional community surveys. There is a pressing need for high-quality evidence on the diagnostic accuracy of epilepsy, the consequences of misdiagnosis, and the psychosocial burden of a new (first-in-a-lifetime) diagnosis of epilepsy. In addition, post-stroke seizures, which may progress to epilepsy, have not been well-studied, especially in people who have received thrombolysis after an acute ischaemic stroke (AIS).
In order to quantify epilepsy misdiagnosis rates, identify reasons and consequences, and to quantify “proportions” and “predictors” of people with seizures who drive and have traffic accidents, I conducted three systematic reviews (including one containing a meta-analysis) of observational studies (published before March 2019). In order to predict who is prone to adverse psychosocial sequelae (i.e. anxiety, depression, behaviour problems, not returning to driving and disability) following a diagnosis of epilepsy, I developed multivariate logistic and linear regression models and ordered logit models using data from the Sydney Epilepsy Incidence Study to Measure Illness Consequences (SEISMIC). In order to determine the frequency, determinants, and significance of early seizures after thrombolysis for people with AIS, I developed logistic regression models using data from the Enhanced Control of Hypertension and Thrombolysis Stroke Study (ENCHANTED).
A false positive diagnosis of epilepsy is common, with syncope and psychogenic non-epileptic paroxysmal events being the most common imitators of seizure where misdiagnosis leads to mismanagement with antiepileptic drugs (AEDs), driving cessation and employment restrictions. The proportion of people with seizures who drive or hold a driver’s license varies widely across studies (from 3% to 98%), and the proportion who have traffic accidents also varies widely (from 0% to 61%). Data were not pooled due to significant heterogeneity between studies. Six variables were consistently associated with driving: male sex, being in paid work, married, older age at seizure onset or at diagnosis, less frequent seizures and taking no or a single AED, while less frequent seizures was protective against traffic accidents. Having a history of psychiatric problems, a higher disability score, lower family functioning and education, encountering economic hardship and perceiving stigma, and taking multiple AEDs independently contributed to psychological problems (i.e. anxiety, depression, behavioural and emotional problems), which occurred in one-fifth to a half of people in the first year after a diagnosis of epilepsy. Early seizures after AIS were associated with male sex, severe neurologic impairment and fever, and although not common, predicted poor recovery at 90 days.
Clinicians need to continue to be cautious when making a diagnosis of epilepsy, and the record of an epilepsy diagnosis in academic research should be validated against standard criteria, and ideally with central specialist adjudication. Further high-quality observational studies are needed to predict those most prone to psychosocial problems, and to produce precise estimations of adverse outcomes from epilepsy. Interventional studies to test the effectiveness of reducing modifiable stressors (adjusted for unmodifiable determinates) are needed to help with the clinical management of people with epilepsy and those with post-stroke seizures.