Choosing treatment wisely

Reflex prescribing

Every prescription deserves a reason and a review.

Different seizures need different treatment plans. A medicine started during an emergency may need a fresh look later. If events continue or side effects become troublesome, the diagnosis and treatment plan deserve another look.

Questions for your next visit Understand reflex prescribing

When a prescription becomes a habit

Here, reflex prescribing means starting or continuing a medicine mainly out of habit, without checking carefully enough why that person needs it and whether it still suits them.

A prescription can be carried forward from one hospital stay or appointment to the next. The original reason may become unclear. Side effects may be overlooked. Continuing events may lead to more treatment before the diagnosis is revisited.

A useful treatment conversation answers three questions:

THE REASON

Why this medicine?

THE PERSON

Why is it right for me?

THE PLAN

When will we review it?

An emergency prescription needs a follow-up plan

In an emergency, doctors may need to treat a seizure before all the answers are available. Once the immediate problem has been treated, the longer-term plan can be considered.

1
Treat the urgent problem
The immediate priority may be stopping a prolonged seizure or treating seizures during a serious illness.
2
Understand what happened
Was it an epileptic seizure? Was it linked to an acute illness or injury? If epilepsy is present, what type is it?
3
Agree on the next review
Does the medicine still have a clear purpose? Is it helping? Who will review the benefit, side effects and need to continue?

Some people need ongoing treatment. Others need a different plan after assessment. A review may confirm that the current medicine remains the best choice.

Where does levetiracetam fit?

Levetiracetam, often known as Keppra, is widely used for epilepsy and in acute seizure care. The best ongoing medicine depends on the type of epilepsy and the person taking it.

The SANAD II trial found that lamotrigine performed better overall as a planned first treatment for newly diagnosed focal epilepsy. This is useful information for a treatment discussion.

Current NICE and WHO guidance still include both lamotrigine and levetiracetam as first options for focal seizures. Your other health conditions, medicines, previous response, side effects and pregnancy considerations can affect the choice.

The epilepsy type matters

Focal epilepsy is common in adults. A whole-body convulsion can occur in focal or generalised epilepsy, so its appearance alone does not settle the diagnosis. Understanding the epilepsy type helps guide treatment; some medicines can worsen particular seizure types when the match is wrong.

What do the prescribing studies show?

Levetiracetam is a frequent choice in many settings. These examples explain why treatment decisions deserve attention. They do not tell us how many prescriptions were inappropriate.

60%

At emergency discharge

135 of 225 seizure patients were taking levetiracetam at discharge from one Spanish emergency department.

2014–2015; age over 15. Includes existing treatment and new prescriptions; prolonged status seizures were excluded.

93%

In a selected hospital group

497 of 532 patients receiving ongoing seizure medication during admission received levetiracetam across five US centres.

2021; adults monitored for acute seizure concerns, with no previous epilepsy. This group received ongoing seizure medicine for at least 48 hours. Included ICU and ward patients; some received several medicines.

57%

First medicine for focal epilepsy

254 of 443 people with newly diagnosed focal epilepsy started levetiracetam as their only seizure medicine.

International research cohort, 2012–2017. Mainly adults, with adolescents included. Recruitment was before SANAD II was published.

80%

Focal epilepsy in an ED study

About 80% of 471 people with confirmed epilepsy in two German emergency departments had focal epilepsy.

Berlin, 2014; adults. This describes epilepsy type, not prescribing, and does not include everyone arriving with a seizure.

Different countries, years and patient groups were studied. The first three figures measure medicine use; the fourth describes epilepsy type. None is a worldwide rate of reflex prescribing, and none measures levetiracetam as a share of every medicine prescribed.

When events continue, revisit the whole picture

Continuing attacks deserve a careful review. The events may need a clearer diagnosis, the epilepsy type may need reassessment, or the treatment may be difficult to take or tolerate. Sometimes an underlying condition needs its own treatment.

Some people have apparent treatment resistance: events continue because of a potentially correctable problem. Others have genuinely drug-resistant epilepsy.

Drug-resistant epilepsy means seizures continue despite two appropriately chosen, adequately tried and tolerated treatment plans. Simply having been prescribed two medicines does not always establish this.

What can go wrong when the diagnosis or treatment does not fit?

Side effects can become an unnecessary burden

Mood, behaviour and other treatment effects deserve attention alongside seizure control. Side effects may be especially difficult to justify if the original reason for treatment is no longer clear.

A review can consider whether symptoms are related to the medicine, the condition being treated or another cause.

Understand side effects

The wrong medicine can worsen some seizure types

Published studies describe people with generalised epilepsy whose seizures worsened after medicines unsuitable for their seizure types, including carbamazepine or phenytoin. This can make epilepsy look harder to control.

These studies support checking the epilepsy classification. They do not show that levetiracetam commonly worsens focal epilepsy.

Some seizure-like events need a different diagnosis and care

Functional seizures are real, involuntary events that can resemble epileptic seizures. They need appropriate care and can occur alongside epilepsy. Reports describe people receiving repeated medicines or intensive emergency treatment because functional seizures were mistaken for epilepsy.

A specialist may use the history, witness accounts and sometimes video-EEG to clarify events. A normal routine EEG on its own does not rule out epilepsy.

Understand other possible causes

An underlying cause may need its own treatment

Some seizures are part of a condition that needs treatment beyond seizure medicines. One example is a particular immune-related brain illness, LGI1 encephalitis.

In an observational study, earlier immune treatment was associated with better outcomes. This illustrates why recognising the cause matters; it does not mean everyone with continuing seizures needs immune treatment.

Can diagnostic errors lead to unnecessary procedures?

Yes. One study found that 13 of 60 people who already had a vagus nerve stimulator and were referred to a specialist monitoring unit had exclusively functional seizures. Their implants had been unnecessary. This was a selected referral group, not a sample of all people receiving VNS.

A small report also described an adult whose seizures did not improve after a brain resection, following misleading findings about the epilepsy type and where seizures began. An unsuccessful operation alone does not show that it was unreasonable to consider surgery. These reports cannot establish how often surgery is unnecessary.

The evidence does not establish that levetiracetam causes permanent drug resistance and then unnecessary surgery. People with genuine drug-resistant epilepsy should still have timely specialist assessment. A surgical evaluation is a thorough review of the diagnosis and treatment options; it does not automatically mean having an operation.

Take this conversation to your next visit

Six questions worth asking

What type of seizures or epilepsy do I have, and how certain is the diagnosis?
Why was this medicine chosen for me?
Was it started for a temporary problem or for ongoing epilepsy?
What is the plan if events continue?
Could changes in my mood, behaviour or daily functioning be related to treatment?
When is my next medication review, and who will review the plan?

Bring your medicine list or photographs of the packs, a record of recent events and side effects, and any witness account or video obtained safely.

Open the appointment toolkit

For health professionals

The practical task is to keep the prescription connected to a diagnosis, a treatment goal and a review plan.

A brief review before treatment is carried forward

Record the indication: established epilepsy, an acute symptomatic seizure, status treatment or prophylaxis.
Check the event and epilepsy classification: preserve uncertainty when the diagnosis is unresolved.
Explain the choice: suitability for this patient, expected benefit and important adverse effects.
Set the review: who will reassess the need for treatment, and when?
Reconstruct previous trials: indication, dose, duration, use, response and reason for stopping before confirming drug resistance.
Refer when appropriate: diagnostic uncertainty and genuine drug resistance both deserve timely specialist assessment.

SANAD II supports lamotrigine for planned initial focal-epilepsy treatment at group level. An adequate levetiracetam trial can still count towards drug resistance. High prescribing share alone is not a measure of inappropriate care.

Explore the evidence

The main message is practical: understand the diagnosis, choose treatment for the individual, and review the plan. The studies and guidance behind this page are available below.

Sources and further reading

This is an educational summary of a targeted evidence review. Prescribing studies used different populations and definitions. Selected reports of harm cannot tell us everyone’s risk, and the research does not quantify a global rate of inappropriate prescribing.

Emergency prescribing in Spain (opens in a new tab)
Toledo et al. Seizure, 2018. A selected emergency-discharge cohort from 2014–2015.
Treatment continued after intensive care (opens in a new tab)
Vorderwülbecke et al. Journal of Neurology, 2018. Seven adult ICUs in Germany.
Acute hospital prescribing in the US (opens in a new tab)
Zafar et al. JAMA Neurology, 2024. Five-centre study of patients monitored for acute seizure concerns in 2021.
Treatment after an acute symptomatic seizure (opens in a new tab)
Herzig-Nichtweiß et al. Annals of Intensive Care, 2023. The observational PROSE study; status epilepticus was excluded.
First medicine for focal epilepsy (opens in a new tab)
Fox et al. Epilepsia, 2023. Human Epilepsy Project prescribing study; recruitment preceded SANAD II.
Epilepsy types seen in emergency care (opens in a new tab)
Reinecke et al. Journal of Neurology, 2022. Adult presentations at two Berlin emergency departments in 2014.
SANAD II: planned treatment for focal epilepsy (opens in a new tab)
Marson et al. The Lancet, 2021. Randomised comparison of lamotrigine, levetiracetam and zonisamide.
NICE treatment guidance (opens in a new tab)
NG217, recommendation 5.2.1: lamotrigine or levetiracetam as first-line monotherapy for focal seizures.
WHO treatment guidance (opens in a new tab)
mhGAP recommendation, 2023. Antiseizure medicines for adults and children.
Emergency treatment of prolonged seizures (opens in a new tab)
Chamberlain et al. The Lancet, 2020. ESETT compared emergency treatments for convulsive status after benzodiazepines.
What drug-resistant epilepsy means (opens in a new tab)
Kwan et al. Epilepsia, 2010. ILAE consensus definition requiring two tolerated, appropriately chosen and used treatment schedules.
When the wrong medicine worsens seizures (opens in a new tab)
Genton et al. Neurology, 2000. Selected patients with juvenile myoclonic epilepsy exposed to carbamazepine or phenytoin; not a levetiracetam study.
Functional seizures treated as status (opens in a new tab)
Jungilligens et al. JNNP, 2021. Reanalysis of RAMPART and ESETT, including children and adults.
Misdiagnosis and apparent resistance (opens in a new tab)
Smith et al. QJM, 1999. Historical specialist-referral cohort; not a population prevalence estimate.
Unnecessary VNS after misdiagnosis (opens in a new tab)
Arain et al. Epilepsy & Behavior, 2011. Selected patients already implanted with a stimulator and referred to one monitoring unit.
Misleading findings before surgery (opens in a new tab)
Fauser et al. Epilepsia Open, 2020. Small, selected series of generalised epilepsies with misleading frontal lesions; not a surgical complication rate.
Timely specialist and surgical evaluation (opens in a new tab)
Jehi et al. Epilepsia, 2022. ILAE expert consensus on referral for epilepsy surgery evaluation.
Recognising a treatable immune cause (opens in a new tab)
Thompson et al. Brain, 2018. Observational study of LGI1-associated seizures and the timing of immunotherapy.
Diagnosis and the limits of a routine EEG (opens in a new tab)
NICE NG217, diagnosis and assessment. A normal routine EEG must not be used to exclude epilepsy.

Evidence reviewed: 18 October 2026. This page supports an individual treatment conversation; it cannot determine whether your medicine should be changed.