Tests for drug-resistant epilepsy: confirm the events, investigate the cause, and define treatment options

Investigation of difficult-to-control seizures usually addresses three questions: whether the recurring events are epileptic, why treatment has not controlled them, and whether additional evidence could change the available options. Different tests provide different kinds of information. They need to be interpreted together. Ask what uncertainty each proposed investigation is intended to resolve rather than selecting the largest or most expensive testing package. NICE NG217: Diagnosis and assessment of epilepsyJehi et al.: Timing of referral for epilepsy surgery evaluation, ILAE consensus, 2022

Key takeaways

These excerpts come from the original article. Read the full sections below for context.

The clinician needs age at first event, frequency, sleep relationship, and the changes before, during, and after an episode. The first observation, progression, injuries, and recovery time can influence interpretation. Broad descriptions such as major and minor fits lose information that may matter to classification. ILAE: Updated classification of epileptic seizures, 2025
During some presurgical assessments, PET or SPECT can add metabolic or perfusion information when history, EEG, and structural imaging leave an unresolved question. The EANM's 2024 epilepsy guideline addresses appropriate use and technical requirements. These investigations are not universal standalone screening tests for everyone with possible epilepsy. Traub-Weidinger et al.: EANM PET and SPECT practice guideline for epilepsy, 2024
Ask the clinician to distinguish confirmed, supported, and uncertain findings and explain which evidence changed management. Not identifying a definite cause does not mean that treatment and safety planning have no useful next steps. Preserve discrepant original results so that they can be reconsidered rather than discarded.

Quick answer

Investigation of difficult-to-control seizures usually addresses three questions: whether the recurring events are epileptic, why treatment has not controlled them, and whether additional evidence could change the available options. Different tests provide different kinds of information. They need to be interpreted together. Ask what uncertainty each proposed investigation is intended to resolve rather than selecting the largest or most expensive testing package. NICE NG217: Diagnosis and assessment of epilepsy Jehi et al.: Timing of referral for epilepsy surgery evaluation, ILAE consensus, 2022

Full guide

Investigation of difficult-to-control seizures usually addresses three questions: whether the recurring events are epileptic, why treatment has not controlled them, and whether additional evidence could change the available options. Different tests provide different kinds of information. They need to be interpreted together. Ask what uncertainty each proposed investigation is intended to resolve rather than selecting the largest or most expensive testing package. NICE NG217: Diagnosis and assessment of epilepsy Jehi et al.: Timing of referral for epilepsy surgery evaluation, ILAE consensus, 2022

Start with a description of the events

The clinician needs age at first event, frequency, sleep relationship, and the changes before, during, and after an episode. The first observation, progression, injuries, and recovery time can influence interpretation. Broad descriptions such as major and minor fits lose information that may matter to classification. ILAE: Updated classification of epileptic seizures, 2025

If there are several patterns, call them event one and event two and keep separate records. Include dates, duration, actual medication use, and relevant context. Preserve both patient and witness accounts when they differ. A diary does not need an elaborate format; it needs to distinguish what was observed from what someone inferred.

Review the medication evidence behind the drug-resistant label

Bring formulations, actual daily use, duration, benefit, and reasons for discontinuing previous medicines. The clinician must establish whether at least two suitable, tolerated, adequately and correctly used schedules failed to achieve sustained seizure freedom. The number of hospitals visited or medicine boxes purchased cannot establish this by itself. NICE NG217: Terms used, drug-resistant epilepsy definition, updated January 2025 ILAE/AES TASK1 report: Revisiting drug-resistant epilepsy, 2023

Missed doses, vomiting, access problems, or independent reductions are relevant clinical information. Their inclusion is not an admission of blame. Some apparent treatment problems require a practical solution, whereas others strengthen the case for timely evaluation of surgery or another approach. An accurate history can prevent repetition of a trial already shown to be ineffective or intolerable.

Understand the limits of a routine EEG

An EEG records electrical activity during the recording period and may support classification. Because relevant abnormalities can be intermittent, a routine study without epileptiform discharges does not exclude epilepsy. Conversely, an abnormality on a report needs clinical interpretation; it should not be equated automatically with a seizure occurring in daily life. NICE NG217: Diagnosis and assessment of epilepsy

Depending on the question, the clinician may consider a sleep recording, ambulatory study, or longer video monitoring. Follow the arranging team's instructions about sleep and medication. Do not independently deprive yourself of sleep, drink alcohol, or stop treatment in an attempt to produce a more informative recording. Such actions can create unnecessary risk without resolving the diagnostic problem.

Video-EEG connects a recorded event with electrical information

Long-term video-EEG can help characterize typical events, distinguish different patterns, and contribute localization evidence in selected patients. Inpatient monitoring requires observation and appropriate response to seizures. Any planned medication adjustment belongs to the monitoring team's protocol and individualized assessment. Tatum et al.: ILAE/IFCN minimum standards for inpatient long-term video-EEG, Epilepsia 2022

Failure to record the usual event during one admission does not resolve every uncertainty. Interpretation depends on the baseline frequency, recording conditions, duration, and information already available. Capturing one event type also does not automatically explain every other episode. A witness should tell the team whether the recorded event matches what usually happens.

Ask what a conclusive or inconclusive recording would mean for the next step. This helps the patient understand why the length of a stay may change or why additional records may be requested. A monitoring admission should have a diagnostic objective beyond simply accumulating hours of EEG.

Functional seizures require explanation and appropriate care

Functional seizures are real health events that can resemble epilepsy and may coexist with it. The AAN's December 2025 guideline emphasizes patient and witness history, available videos, and recording typical events with video-EEG where feasible to increase diagnostic certainty. Assessment also considers coexisting conditions. AAN: Management of Functional Seizures, clinical guideline summary, December 2025

If this diagnosis is proposed, request an explanation of the supporting evidence, remaining uncertainty, and treatment plan. It should not be assigned casually because tests are normal, and it does not mean that a person is pretending. Medication decisions depend on whether epilepsy or another indication is also present. Do not abruptly discontinue antiseizure treatment merely because a new diagnostic term has been discussed.

ECG and blood tests may investigate a different explanation

Some transient loss of consciousness has a cardiac cause, making an ECG and further cardiac assessment relevant in appropriate circumstances. Glucose, electrolyte, and systemic disturbances can also cause or provoke events. The clinician selects investigations from the history and current findings; not every convulsion is solved by increasing an antiseizure medicine. NICE NG217: Diagnosis and assessment of epilepsy

In a treated patient, kidney and liver function, blood counts, or selected medication levels may address safety and exposure. Drug concentrations are not mandatory at every visit for everyone, and a level within a laboratory range does not guarantee seizure control. Record the timing of the blood sample relative to the last dose. NICE NG217: Full guideline, updated January 30, 2025

When a test is abnormal, ask whether it may explain the events, result from treatment, or represent a separate issue. Those interpretations lead to different actions. The laboratory flag alone does not tell the patient which medicine to change.

An epilepsy MRI depends on acquisition and interpretation

Structural imaging for epilepsy uses suitable sequences and resolution, together with experienced reading. The ILAE HARNESS recommendations aim to standardize acquisition and improve identification of subtle abnormalities. Whether a scan needs repeating depends on its quality, timing, and the current clinical question, not simply on arrival at a new hospital. Bernasconi et al.: ILAE structural MRI recommendations and HARNESS protocol, 2019

A report stating that no obvious abnormality was seen means that the available study did not show a clear lesion. It does not independently exclude epilepsy or remove every surgical evaluation option. Conversely, a visible lesion requires correlation with the clinical and electrical evidence before its role is established.

Provide details of metal implants, stimulators, and previous surgery before booking. The imaging team needs the specific device information to check the applicable scanning conditions. Having undergone an MRI previously is not sufficient evidence that every later scan under different conditions is safe.

PET and SPECT answer selected localization questions

During some presurgical assessments, PET or SPECT can add metabolic or perfusion information when history, EEG, and structural imaging leave an unresolved question. The EANM's 2024 epilepsy guideline addresses appropriate use and technical requirements. These investigations are not universal standalone screening tests for everyone with possible epilepsy. Traub-Weidinger et al.: EANM PET and SPECT practice guideline for epilepsy, 2024

Results can depend on the relationship to seizures, injection, and acquisition. A region of reduced metabolism may extend beyond the tissue ultimately relevant to treatment. A colored image therefore should not independently determine a surgical boundary. Discuss radiation, pregnancy, breastfeeding, or sedation-related considerations with the team when applicable.

If another hospital has already performed the examination, ask whether its original data can be reviewed. Repeating it may sometimes be justified, but the reason should be explained. The clinical value lies in how the result changes the overall hypothesis, not simply in the technology's name.

Neuropsychological assessment does more than measure intelligence

Testing can document memory, language, attention, emotional, and other functions, establish a pretreatment baseline, and inform discussion of cognitive risk after an intervention. It contributes to the whole assessment; a single score should not by itself determine whether a patient can have surgery. Baxendale et al.: ILAE neuropsychological assessment in epilepsy surgery, 2019

Language, education, sleep, mood, and medication state can affect performance. International patients should identify their first and routinely used languages before assessment in China. Ask about suitable testing and interpretation. Informal word-for-word translation does not necessarily reproduce a validated test, and the team needs to recognize limitations before attributing difficulty to brain dysfunction.

Explain practical concerns as well as scores. A person who manages complex verbal tasks for work may have different priorities from someone mainly concerned about everyday memory. The assessment can help make potential benefits and risks relevant to those needs.

Consider genetic testing when the phenotype supports its value

Early onset, developmental features, family clustering, or other findings may increase the usefulness of a genetic assessment. Adults whose epilepsy began in childhood before testing was widely available may still merit consideration. The method and scope should follow the clinical phenotype and how an informative result might influence care. Krey et al.: Current practice in diagnostic genetic testing of the epilepsies, ILAE Genetics Commission, 2022

Discuss positive, negative, and uncertain outcomes before testing, including family implications. A negative result does not prove that no genetic cause exists, while a detected variant does not automatically identify a targeted treatment. Reanalysis may become relevant as knowledge changes and should be coordinated with the appropriate epilepsy and genetics services.

Keep the full original report, including method and limitations. A shortened summary saying genetic epilepsy can lose the distinction between a confirmed cause and a candidate finding. That distinction matters when later clinicians reconsider treatment.

Immune, infectious, and metabolic investigations need a clinical reason

Rapidly increasing seizures accompanied by new memory, behavioral, psychiatric, or consciousness changes may require timely investigation for encephalitis and other causes. Assessment can involve blood, cerebrospinal fluid, imaging, and EEG. A broad antibody panel alone cannot establish the entire diagnosis, and alternative explanations such as infection must be considered. Abboud et al.: Autoimmune encephalitis diagnosis and acute management best-practice recommendations, 2021

Longstanding drug resistance does not mean that every patient needs an identical antibody panel. Positive or negative findings require interpretation in relation to the phenotype, specimen, and assay. Do not independently start steroids or immune treatment for a borderline result. Rapid deterioration needs local assessment rather than waiting for a planned consultation abroad.

Invasive EEG is reserved for a question that justifies its risks

SEEG and other intracranial recordings address particular localization uncertainties remaining after noninvasive work. Electrode implantation carries risks including bleeding and infection. It is not an upgraded routine checkup. The team should first identify the hypotheses being tested and explain how the findings could influence treatment choices. NINDS: Epilepsy and Seizures Jehi et al.: Timing of referral for epilepsy surgery evaluation, ILAE consensus, 2022

Not all patients with drug resistance need invasive recording, and implantation does not guarantee that a resection or ablation will follow. Ask where current evidence agrees, where it conflicts, and what the proposed recording is intended to distinguish. Also discuss the pathway if the result does not support a sufficiently safe intervention.

Organize testing in China through the receiving clinical service

Submit original EEG and imaging, the medication history, and event descriptions through the actual department. CAAE center information can help locate services, but confirm the campus, relevant adult or pediatric expertise, monitoring facilities, and multidisciplinary assessment required. Several departments may perform tests; identify who will integrate their conclusions. 中国抗癫痫协会：第五批一、二级癫痫中心评审结果，2025年9月28日中国抗癫痫协会官方网站及CAAE癫痫地图入口

State the usual event frequency, caregiver needs, language, and constraints on the stay. These details allow the hospital to explain the scope of a realistic assessment. No service should assume it can guarantee a captured event or a clear lesion within a fixed number of days. Ask for costs of the actual indicated work rather than treating an undefined complete package as a medical recommendation.

Leave with both the conclusions and the unanswered questions

Ask the clinician to distinguish confirmed, supported, and uncertain findings and explain which evidence changed management. Not identifying a definite cause does not mean that treatment and safety planning have no useful next steps. Preserve discrepant original results so that they can be reconsidered rather than discarded.

Obtain formal reports and available source data, record the conditions for another review, and identify whom to contact if events change. Give the receiving home clinician the conclusions and any medication changes. The result of an investigation should be a clearer next decision, not merely a larger collection of papers.