Considerations in the diagnosis of epilepsy I. Distinguish between evoked and nonevoked seizures Not all seizures should be diagnosed as epilepsy. By definition, a patient’s seizures must be nonprovoked to be diagnosed as epilepsy, and evoked seizures, even if recurrent, are usually not considered for a diagnosis of epilepsy. The misdiagnosis of recurrent acute symptomatic seizures as “symptomatic epilepsy” inevitably leads to overdiagnosis and treatment, as well as to unreliable epidemiological findings in epilepsy. Conditions in which seizures are present but not usually diagnosed as epilepsy include: benign neonatal seizures, febrile convulsions, alcohol or drug withdrawal seizures, and seizures that occur during the acute phase of a central nervous system or systemic disease. Recurrent reflex seizures can be diagnosed as epilepsy according to the 2014 ILAE Clinical Utility Definition of Epilepsy, even if each seizure appears to be “triggered” (Appendix 1). The role of history and ancillary tests in the diagnosis of epilepsy The history is the most important basis for the diagnosis of epilepsy, which is largely a clinical diagnosis. By definition, epilepsy is diagnosed when two non-induced seizures are clinically present, and medication can usually be considered. In most cases, a detailed history, especially a history of seizures, is sufficient to determine whether the seizure symptoms are epileptic, or even to make a preliminary diagnosis of seizure type and type of epilepsy (syndrome), with later EEG and imaging often serving as a means to further verify or clarify the prior diagnosis. An abnormal EEG does not necessarily lead to a diagnosis of epilepsy, and a normal EEG does not exclude epilepsy. Patients who have had several typical grand mal seizures in a short period of time, but fail to diagnose epilepsy and delay treatment because of a normal EEG should be avoided. The most common reason for misdiagnosis of epilepsy is inadequate history taking. Seizures are often brief, and it is unlikely that the physician will witness a seizure, so a detailed and well-organized history is especially important. The patient should be seen with a witness to the seizure in order to obtain a complete history. When the patient’s description is unclear at the time of the visit, it is necessary for the physician to interview the seizure witness by telephone. If possible, it is recommended that the patient or family member videotape the seizure on a cell phone or home video camera for the physician to analyze during the visit. Also, in cases where the patient or witness is unclear, it is good to have them watch a variety of videos of typical seizures, which often allows them to find the seizure that most closely resembles the patient’s presentation. If it is difficult to obtain a reliable history, explain the importance of the history to the patient so that he or she can be observed in the event of another seizure and provided at a follow-up visit. Avoid missing “minor seizures” Complete information on the type of seizure is important for the diagnosis of the type of epilepsy (syndrome). When taking a history, it is important to focus on both obvious seizures (e.g., grand mal seizures) and certain “minor seizures” that are often overlooked or not actively reported by patients or witnesses of seizures, such as aura seizures, myoclonic seizures, and focal seizures with minimal impairment of consciousness. For example, in an adolescent patient who complains of several primary grand mal seizures and has a normal past history, if the history asks for “shaking” of the limbs after waking up in the morning, which is often overlooked by the patient, the clinical consideration is “juvenile myoclonic epilepsy”, otherwise Otherwise, it may be considered “generalized epilepsy with grand mal seizures only”. By definition, the “gold standard” for seizure diagnosis is to establish a “causal relationship” between abnormal EEG activity and clinical manifestations during the seizure phase, which can be achieved by long-range video-EEG monitoring. Of course, long-range monitoring is neither practical nor necessary in all patients. In those cases where the nature of the seizure is not clear from a detailed history, long-range video-EEG monitoring can be performed to clarify the diagnosis. In addition, the above-mentioned “gold standard” should be used for the diagnosis of “abdominal epilepsy”, “headache epilepsy”, and “epilepsy with xxx as the only seizure manifestation”, which are easily diagnosed in China. The “gold standard” should be measured and tested. Of course, the limitations and shortcomings of long-range video-EEG monitoring should also be understood in practice. Before diagnosing drug-refractory epilepsy, care should be taken to rule out “pseudo” drug-refractory epilepsy. The following should be considered: (1) non-epileptic seizures; (2) misclassification of seizures (e.g., misdiagnosis of aphasic seizures as complex partial seizures); (3) inappropriate drug selection for seizure type (e.g., carbamazepine to control aphasic seizures); (4) inadequate drug doses or inappropriate drug administration methods; (5) poor patient compliance; (6) controllable triggers that aggravate seizures (e.g., excessive alcohol consumption, lack of sleep, etc.) lack of sleep, etc.) (7) other etiologies that can lead to epilepsy refractory to treatment (e.g., vitamin B6 dependence, glucose transporter I deficiency, etc.). In addition, some patients with epilepsy may have both seizures and non-epileptic seizures, which should be differentiated, and if necessary, long-range video-EEG monitoring should be performed to clarify the diagnosis. Avoid increasing the dose of medication or changing medication frequently to control “intractable epilepsy” because the seizure symptoms are mistaken for seizures. Although the diagnosis and treatment of epilepsy are closely related, they are not necessarily linked. The diagnosis of epilepsy does not always require treatment. For example, patients with benign partial epilepsy in children with sparse seizures or epilepsy with mild seizures (e.g., aura-only seizures) may choose not to treat them. The decision to treat a patient with epilepsy depends on a variety of factors, including the patient’s wishes and the individualized benefit-risk ratio of taking/not taking medication. On the other hand, initiation of treatment may be considered without a diagnosis of epilepsy. For example, in patients with recurrent seizures presenting in the acute phase of encephalitis, medication is usually administered clinically despite the absence of a diagnosis of epilepsy.