This page shows what was measured, who it was measured in, and what that does not settle.
What Carbamazepine does in the body
Epilepsy, facial nerve pain and bipolar mania.
Nerve cells fire by letting sodium rush in through pores in their outer membrane. After each firing a pore spends a moment shut and unavailable, and carbamazepine binds to it in exactly that state and holds it there. A cell firing at a normal rate barely notices. A cell firing over and over, as it does in a seizure or in a trigeminal pain attack, finds more and more of its pores locked out, so the burst starves itself. The drug also switches on the liver enzymes that destroy it, so it speeds up its own removal over the first month.
What happened in people
It matched or beat older seizure medicines, but people stopped it sooner than lamotrigine.
✓ Reviewed first-read answer
Where this came from
A person wrote this and a reviewer approved it against this exact record. It carries no effect size.
A reviewer approved this against this record. The reviewed-claim record carrying the exact population and effect size does not exist yet.
No source is stored against this line.
The limit that matters most
Its official label still says its biological action remains unknown.
Where it acts
Axonal membrane of cortical neurons and of the trigeminal nerve root (voltage-gated sodium channels)
Kind of result
The kind of result is not recorded
Supervision
Professional supervision is normally required. A regulator classifies this substance in a way that restricts how it is supplied.
What the registries record it as
The substance registry classes this as chemical.
FDA substance registry · 33CM23913M · read 2026-08-29
Where each sentence above came from
No recorded sentence describes the change this makes in a way that stands on its own. The page opens with what it is taken for instead, and the recorded explanation follows below.
Shown as the opening line on this page.
The limit a reviewer approved as the one that matters most here.
The four opening statements run to 124 words.
Words this page uses
Four words worth knowing first
Chosen from what this page shows, with each one explained before the word it depends on.
Comparator
A comparator is whatever the treatment was measured against.
A picture of it, and where the picture fails
It is like the other runner in a race.
Where that stops being true. A race has one winner. A study can show both arms improved.
What people get wrong. Results are read without asking what the other group got. Beating nothing is not beating a treatment.
The control condition against which the experimental intervention is assessed.
Randomisation
Randomisation means chance decides who gets which treatment.
A picture of it, and where the picture fails
It is like a coin toss deciding the groups.
Where that stops being true. A coin toss is fair once. Fairness here comes from doing it for everyone.
What people get wrong. It is thought to make groups identical. It makes them similar on average, including on things nobody measured.
Allocation of participants to arms by a chance process, so that unmeasured factors are balanced in expectation.
Absolute difference
An absolute difference is how many more people in a hundred were affected.
A picture of it, and where the picture fails
It is like counting heads in two rooms of a hundred.
Where that stops being true. Heads are easy to count. Study results carry a margin of error too.
What people get wrong. It is confused with a percentage change, which can look far larger.
The arithmetic difference in event rates between arms.
Confidence interval
A confidence interval is the range the true answer is likely to sit in.
A picture of it, and where the picture fails
It is like a weather forecast giving a range rather than one number.
Where that stops being true. A forecast range covers tomorrow. This one covers what the study could resolve.
What people get wrong. The middle number is read as the answer. A range crossing zero means no change is still possible.
An interval estimate that would contain the true parameter in a stated proportion of repeated studies.
What happened in people◇Read from sources, not yet reviewed
What happened in people
Each card is one study: the exact question it asked, who was in it, and whether it showed what it set out to show.
Overall treatment success, combining seizure control and tolerability, across carbamazepine, phenobarbital, phenytoin and primidone
Highest with carbamazepine or phenytoin, lowest with primidone, P<0.002; complete control of partial seizures better than primidone or phenobarbital, P<0.03
Repeated elsewhere
Replicated
What this does not prove. Meeting one endpoint in one population does not carry to other goals or other people.
The limits of this study, and where it came from
Main limit. No limitation is recorded for this study.
How far it carries. This study is recorded from the curated record, not from the registry match.
Form and route. Oral tablet, chewable tablet, extended-release tablet, extended-release capsule and oral suspension
Interval reported. Not recorded in this summary
Written into the record, not signed off as a reviewed claim.
Seizure control and composite control-plus-toxicity score, carbamazepine versus divalproex sodium
✓ The study showed what it set out to show
Who was studied
VA Cooperative Study 264 (Mattson 1992)
How many people
480
Study design
Double-blind randomised comparative trial, 1 to 5 years of follow-up
Compared against
Not recorded for this study
Kind of result
A number that stands in for health
What was found
Complex partial seizures per month 0.9 versus 2.2, P=0.01; composite score favoured carbamazepine, P<0.001; no difference for secondarily generalised tonic-clonic seizures
Repeated elsewhere
Replicated
What this does not prove. Meeting one endpoint in one population does not carry to other goals or other people.
The limits of this study, and where it came from
Main limit. Rash occurred in 11% on carbamazepine against 1% on valproate (P<0.001), a difference whose genetic basis was not identified for another twelve years.
How far it carries. This study is recorded from the curated record, not from the registry match.
Form and route. Oral tablet, chewable tablet, extended-release tablet, extended-release capsule and oral suspension
Interval reported. Not recorded in this summary
Written into the record, not signed off as a reviewed claim.
Co-primary: time to treatment failure and time to 12-month remission, carbamazepine against gabapentin, lamotrigine, oxcarbazepine and topiramate
✗ The study did not show it
Who was studied
SANAD arm A (ISRCTN38354748)
How many people
1721
Study design
Unblinded randomised controlled trial, five parallel arms
Compared against
Not recorded for this study
Kind of result
A number that stands in for health
What was found
Lamotrigine better than carbamazepine for time to treatment failure, HR 0.78 (95% CI 0.63 to 0.97); for 12-month remission carbamazepine held a non-significant advantage, HR 0.91 (0.77 to 1.09)
Repeated elsewhere
Partially Replicated
What this does not prove. A study that did not show an effect does not show that no effect exists anywhere.
The limits of this study, and where it came from
Main limit. The trial was unblinded, so the treatment-failure endpoint, which depends on a clinician deciding to stop a drug, was open to expectation on both sides.
How far it carries. This study is recorded from the curated record, not from the registry match.
Form and route. Oral tablet, chewable tablet, extended-release tablet, extended-release capsule and oral suspension
Interval reported. 95% CI 0
Written into the record, not signed off as a reviewed claim.
Prospective cohort with historical control, 23 hospitals
Compared against
Not recorded for this study
Kind of result
A number that stands in for health
What was found
Zero cases observed against approximately 10 expected from a historical incidence of 0.23%, P<0.001
Repeated elsewhere
Replicated
What this does not prove. Meeting one endpoint in one population does not carry to other goals or other people.
The limits of this study, and where it came from
Main limit. Mild transient rash still occurred in 4.3% and hospitalising rash in 0.1%. The allele predicts SJS and TEN only; the label states it does not predict maculopapular eruption or DRESS.
How far it carries. This study is recorded from the curated record, not from the registry match.
Form and route. Oral tablet, chewable tablet, extended-release tablet, extended-release capsule and oral suspension
Interval reported. Not recorded in this summary
Written into the record, not signed off as a reviewed claim.
RNAWiki holds 4 written-up human studies for this substance, each with the question it asked.
How we know it
Each card names the study, the number of people and what the study set out to measure.
What this does not prove
These summaries were written by a person and have not been signed off as reviewed claims.
Why it matters
A study that failed is as informative as one that worked, and both are here.
What we still do not know
No reviewed claim exists, so no exact population, effect size or interval is published yet.
What it changes in the body◇Read from sources, not yet reviewed
The path through the body
From what a person takes to what changes. A solid line is a step measured in people. A dashed line is a step nobody has measured that way.
Start
Carbamazepine
What a person takes: Oral tablet, chewable tablet, extended-release tablet, extended-release capsule and oral suspension.
The measurement behind this step
Formulation matters more here than for most drugs. The suspension peaks at about 1.5 hours, the conventional tablet at 4 to 5 and the extended-release tablet at 3 to 12, and the extended-release tablet given twice daily reproduces the steady-state levels of the conventional tablet given four times daily. There is no intravenous carbamazepine in general use, which is one reason it is absent from status epilepticus protocols.
Getting in
Swallowed, absorbed slowly, and increasingly destroyed by the liver it wakes up
Absorption is slow and varies between formulations. Over the first three to five weeks the liver learns to clear the drug faster, so blood levels drift down on an unchanged dose.
──Measured in people. Written by a person from the study named beside the step. Not signed off as a reviewed claim.
The measurement behind this step
Peak plasma levels arrive at about 1.5 hours from suspension, 4 to 5 hours from conventional tablets and 3 to 12 hours from the extended-release tablet. Plasma protein binding is 76%. Metabolism runs mainly through CYP3A4 to carbamazepine-10,11-epoxide, which is itself anticonvulsant in animal models and is cleared by microsomal epoxide hydrolase. Autoinduction of CYP3A4 completes after 3 to 5 weeks, taking the half-life from 25 to 65 hours down to 12 to 17.
It reaches brain tissue at roughly its free concentration in blood
The molecule is small and fat-soluble enough to cross into the brain, and the amount that gets there tracks the fraction not stuck to blood proteins.
╌╌Described, with no measurement named. No measurement is named for this step, so it is drawn as an unconfirmed link.
The measurement behind this step
The CSF to serum ratio is 0.22, closely matching the 24% of carbamazepine that is unbound in serum, which is the expected relationship for passive distribution of a free drug. The relevant compartment is the axonal and presynaptic membrane of cortical neurons, and for trigeminal neuralgia the root entry zone of the fifth cranial nerve.
It binds sodium channels that have just fired, not ones at rest
Each sodium pore passes through a shut, unavailable state right after it opens. Carbamazepine binds that state and holds the pore there, so pores that have just been used are the ones taken out of service.
╌╌Described, with no measurement named. No measurement is named for this step, so it is drawn as an unconfirmed link.
The measurement behind this step
Binding to the inactivated conformation of voltage-gated sodium channel alpha subunits shifts steady-state inactivation to more negative potentials and slows recovery from inactivation. Apparent potency is far higher from a depolarised holding potential than from a hyperpolarised one, which is what makes the block state-dependent rather than simply concentration-dependent.
A cell firing occasionally loses almost nothing. A cell firing in a fast train loses more of its pores with every spike, so the train fades instead of building.
╌╌Described, with no measurement named. No measurement is named for this step, so it is drawn as an unconfirmed link.
The measurement behind this step
Cumulative block across a stimulus train reduces sustained high-frequency firing. The label describes the observable version of this in animals: reduced polysynaptic responses and blocked post-tetanic potentiation, and abolition of pain evoked by infraorbital nerve stimulation. The label then states that the mechanism of action remains unknown.
Fewer focal seizures, and fewer facial pain attacks
The measured results are two: better control of focal seizures than phenobarbital or primidone in a blinded 622-patient trial, and relief of trigeminal neuralgia, which is what the drug was first licensed for in 1968.
╌╌Described, with no measurement named. No measurement is named for this step, so it is drawn as an unconfirmed link.
The measurement behind this step
Efficacy in focal epilepsy is established against three older comparators (VA 118), against valproate (VA 264) and across four newer drugs (SANAD arm A). The extended-release capsule was separately approved for acute manic and mixed episodes of bipolar I disorder in 2004 as Equetro, a different indication reached through different registration trials.
No suggested links are held for this record, so nothing is hidden from this path.
What we know
The record describes 5 steps from taking it to an effect in a person.
How we know it
Each step names the study behind it in the technical detail beside it.
What this does not prove
A change inside the body is a reason to look. It is not a result in a person.
Why it matters
A reader can see exactly where the chain stops being measured in people.
What we still do not know
No reviewed claim binds any of these steps to a named population.
What is missing or unclear◇Read from sources, not yet reviewed
Were people like you studied?
RNAWiki cannot tell whether a study fits you. It can show who was in it, and where the result stops carrying.
Who was studied
People similar to this profile were included.
People with focal epilepsy, people with trigeminal neuralgia, and a smaller group with bipolar I disorder. Prescribing has fallen in high-income countries because of interactions and rash risk, but it remains a WHO essential medicine and a mainstay where newer drugs are unaffordable.
Who is missing from the studies
Which groups were under-represented in the studies has not been recorded for this substance. Pregnancy, breastfeeding, children, older adults, and people with liver or kidney conditions are commonly excluded from trials; nothing here says whether they were.
What the label states about particular groups
On pediatric, the label states: “Bipolar Disorder and Pain of Trigeminal Neuralgia The safety and effectiveness of EQUETRO have not been established in pediatric patients.”
US prescribing information · be478f3c-40f6-47cc-8ab9-f420a9372b1c · read 2026-08-30
On older people, the label states: “Clinical studies of EQUETRO did not include sufficient numbers of subjects age 65 and over to determine whether they respond differently from younger subjects.”
US prescribing information · be478f3c-40f6-47cc-8ab9-f420a9372b1c · read 2026-08-30
On people who are pregnant, the label states: “Pregnancy Exposure Registry There is a pregnancy exposure registry that monitors pregnancy outcomes in women exposed to antiepileptic drugs (AEDs), such as EQUETRO, during pregnancy.”
US prescribing information · be478f3c-40f6-47cc-8ab9-f420a9372b1c · read 2026-08-30
On people who are breastfeeding, the label states: “Carbamazepine and its epoxide metabolite are present in human milk.”
US prescribing information · be478f3c-40f6-47cc-8ab9-f420a9372b1c · read 2026-08-30
Where the result stopped carrying
The drug was first investigated as an antidepressant on its structural resemblance to imipramine, and reached the market for trigeminal neuralgia in 1968 before it reached it for epilepsy in 1974
It lost first-line status in United Kingdom focal epilepsy guidance after SANAD, on tolerability rather than on seizure control
For forty years its severe rash was classified as idiosyncratic and unpredictable, an inference falsified by a one-page report in 2004
This is a scope explorer, not a diagnosis engine.
It shows who was studied so you can see whether people similar to you were included.
RNAWiki cannot tell whether a study fits you. It can show who was in it.
What it would be like to take◇Read from sources, not yet reviewed
Why it might seem to do nothing
A real effect and a noticed effect are different things. These are the recorded reasons the two come apart.
It was studied in different people
The people in the studies were not much like the reader.
On this record: Which groups were under-represented in the studies has not been recorded for this substance. Pregnancy, breastfeeding, children, older adults, and people with liver or kidney conditions are commonly excluded from trials; nothing here says whether they were.
The evidence may simply be wrong
Small early studies often do not hold up.
On this record: RNAWiki has not yet published a reviewed conclusion for this use.
Other reasons RNAWiki checked and found nothing for (11)
It was studied for a different goal
The studies measured something else entirely. Nothing in this record points to this reason.
A different form was studied
The studied form is not the form on the shelf. Nothing in this record points to this reason.
There was nothing to correct
Where a level is already normal, topping it up may change nothing. Nothing in this record points to this reason.
Something else had to happen too
In the studies it was paired with training, a diet or another treatment. Nothing in this record points to this reason.
The change is too small to feel
A real change can still sit below what a person notices. Nothing in this record points to this reason.
Day-to-day swing hides it
Sleep, food, stress and time of day move most numbers more than this would. Nothing in this record points to this reason.
Not enough time yet
The studies ran longer than the person has waited. Nothing in this record points to this reason.
It was not taken as studied
Missed days change the result, and studies count them. Nothing in this record points to this reason.
Something else changed at the same time
Two changes at once cannot be told apart afterwards. Nothing in this record points to this reason.
The product may not be what it says
Contents of a sold product are not always what the label states. Nothing in this record points to this reason.
No recorded reason
RNAWiki has nothing stored that would explain it. Nothing in this record points to this reason.
None of these is a reason to take more. Taking more is not a step this page ever suggests.
What it would be like to take◇Read from sources, not yet reviewed
What taking it involves
The recorded facts about getting hold of it and using it. Nothing here is a suggestion about what you should do.
How it is supplied
Prescription only
❝ Quoted from a source
Where this came from
Copied from a source RNAWiki stored, with the source named.
A person wrote this into the record with the study named beside it. No reviewer has signed it off.
The form and route recorded on this substance.
No source is stored against this line.
Who oversees it
Professional supervision is normally required. A regulator classifies this substance in a way that restricts how it is supplied.
❝ Quoted from a source
Where this came from
Copied from a source RNAWiki stored, with the source named.
Suppression classes recorded: S3, S6.
No source is stored against this line.
What is in the pack
Formulation matters more here than for most drugs.
✎ Source-linked record
Where this came from
A person wrote this into the record with the study named beside it. No reviewer has signed it off.
Recorded notes on how this is sold. The rest of the recorded wording: 5 hours, the conventional tablet at 4 to 5 and the extended-release tablet at 3 to 12, and the extended-release tablet given twice daily reproduces the steady-state levels of the conventional tablet given four times daily. There is no intravenous carbamazepine in general use, which is one reason it is absent from status epilepticus protocols.
No source is stored against this line.
Where it is registered
Regulatory records are listed in the technical disclosure at the foot of this page.
# Counted from records
Where this came from
A count of rows RNAWiki holds. It describes our records, not your body.
Register entries are stored per jurisdiction and shown with their dates.
No source is stored against this line.
Why people stop
Not recorded.
∅ Nothing recorded
Where this came from
RNAWiki holds nothing here and says so rather than guessing.
No record of why people stopped taking it is stored for this substance.
No source is stored against this line.
The recorded stepping schedule
What follows is the schedule printed on the label, recorded as it is written there. It is a record of what was done, not a recommendation.
US prescribing information · 047bc284-060a-4db9-bf37-75d10f95a0a6 · read 2026-08-27
Initial (adults and children over 12 years of age): Either 200 mg twice a day for tablets and XR tablets, or 1 teaspoon four times a day for suspension (400 mg/day) — Label-stated initial dosage for epilepsy
US prescribing information · 047bc284-060a-4db9-bf37-75d10f95a0a6 · read 2026-08-27
At weekly intervals: Add up to 200 mg/day using a twice a day regimen of Tegretol-XR or a three times a day or four times a day regimen of the other formulations until the optimal response is obtained — Label-stated weekly increase for epilepsy
US prescribing information · 047bc284-060a-4db9-bf37-75d10f95a0a6 · read 2026-08-27
Maintenance: Adjust dosage to the minimum effective level, usually 800 to 1200 mg daily — Label-stated maintenance dosage for epilepsy
US prescribing information · 047bc284-060a-4db9-bf37-75d10f95a0a6 · read 2026-08-27
What it would be like to take◇Read from sources, not yet reviewed
What can go wrong
Sorted by where each line came from. A regulator's label and a report somebody sent in are not the same kind of fact.
·Worth asking a clinician aboutWritten into the record from the studies named on this page
The boxed warning has two halves. The first is dermatologic: serious and sometimes fatal SJS and TEN, estimated at 1 to 6 per 10,000 new users in mainly Caucasian populations and about ten times that in some Asian countries, with HLA-B*1502 screening required before treatment in people with ancestry across broad areas of Asia. The second is haematologic: aplastic anaemia and agranulocytosis at 5 to 8 times the general-population rate, on a baseline of roughly six and two cases per million per year. Beyond the box, hyponatraemia is common, the drug is a potent CYP3A4 inducer that lowers the levels of hormonal contraception and many other medicines, and it can worsen absence and myoclonic seizures in generalised epilepsy. The class-wide suicidality warning applies.
Nobody counted how many people took this and were fine, so this cannot be turned into a rate.
Over a long time. No completed tested study window is recorded, so long-term effects are not established by the registry record.
Groups the studies covered thinly. Which groups were under-represented in the studies has not been recorded for this substance. Pregnancy, breastfeeding, children, older adults, and people with liver or kidney conditions are commonly excluded from trials; nothing here says whether they were.
What is missing or unclear◇Read from sources, not yet reviewed
Does the exact form matter?
Evidence belongs to the exact thing that was tested. A different salt, a different mixture or a different preparation is a different question.
A person wrote this into the record with the study named beside it. No reviewer has signed it off.
The form used in the studies cited on this page.
No source is stored against this line.
What is sold
A recorded note compares this form with the others that are sold. It is kept below, word for word.
§ A fixed RNAWiki sentence
Where this came from
Wording RNAWiki always uses, not a finding about this substance.
The recorded note, unchanged: The suspension peaks at about 1.5 hours, the conventional tablet at 4 to 5 and the extended-release tablet at 3 to 12, and the extended-release tablet given twice daily reproduces the steady-state levels of the conventional tablet given four times daily. There is no intravenous carbamazepine in general use, which is one reason it is absent from status epilepticus protocols.
No source is stored against this line.
What is recorded as being sold
175 products list this as an active ingredient in the United States drug directory. 175 of them contain it and nothing else.
FDA National Drug Code directory · 0615-8125 · read 2026-08-29
They are sold as capsule, extended release, powder, suspension, tablet, tablet, chewable and tablet, extended release, taken oral.
FDA National Drug Code directory · 0615-8125 · read 2026-08-29
The regulator's established pharmacologic class for it is cytochrome p450 1a2 inducers [moa], cytochrome p450 2b6 inducers [moa] and cytochrome p450 2c19 inducers [moa].
FDA National Drug Code directory · 0615-8125 · read 2026-08-29
90 published labels name it as an active ingredient. 90 of them describe this substance alone, which is where its own label text on this page comes from.
US prescribing information · 32dc301e-881d-47b0-ad16-f9ff31d19324 · read 2026-08-29
Those labels are classed as human prescription drug.
US prescribing information · 32dc301e-881d-47b0-ad16-f9ff31d19324 · read 2026-08-29
12 marketed supplement labels list this ingredient, classed as non-nutrient/non-botanical, other combinations and vitamin.
Those labels carry all other, nutrient and structure/function claims. A claim of that kind is written by the manufacturer and is not assessed by any regulator, so its presence says nothing about whether it is true.
Recorded price in US: 0.09596 USD per one millilitre, across 4 priced products whose strengths and pack forms differ, as of 2026-08-26, paid by retail pharmacies buying the product, as surveyed by the Centers for Medicare & Medicaid Services. It is not a list price, an insurance price, or what anyone pays at a counter..
Recorded source · 2026-08-26 · read 2026-08-28
Recorded price in US: 0.11904–0.25286 USD per one unit as the pricing file counts it — a tablet, capsule, patch or single item, across 27 priced products whose strengths and pack forms differ, as of 2026-08-26, paid by retail pharmacies buying the product, as surveyed by the Centers for Medicare & Medicaid Services. It is not a list price, an insurance price, or what anyone pays at a counter..
Recorded source · 2026-08-26 · read 2026-08-28
Evidence carries across two names for one substance. It does not carry across a salt, a mirror form or a mixture.
What it would be like to take◇Read from sources, not yet reviewed
What you could measure
This one is decided with a clinician, so this section holds questions to ask rather than a plan to run.
Self-experiment planning is not offered for a prescription or clinician-supervised medicine. The questions below are for a clinician or pharmacist.
Questions worth asking
Which of the trials of Carbamazepine studied people like me?
What was measured, and for how long?
Was the result a laboratory value or a health outcome?
What would we watch for, and when would we stop?
Tracking can show whether something changed for you. It cannot show what caused it.
RNAWiki records evidence. It does not say whether this substance is right for you.
What is missing or unclear◇Read from sources, not yet reviewed
Claims that go past the evidence
Things said about this substance that sound reasonable, and the exact step that is missing between the evidence and the claim.
✗Goes past the evidence
That sodium-channel block is the established mechanism in people: it is the best cellular account, but the label says the mechanism remains unknown and the active-metabolite question is open
Why it sounds right. It follows from something real on this page: a body step, an animal result or a related finding.
The missing step. Nobody measured this in people for this use, or the study that did measure it did not show it.
What would change this
A study in people, for this exact use, measuring this exact thing, that a reviewer signs off.
RNAWiki has not yet published a reviewed conclusion for this use.
✗Goes past the evidence
That periodic blood counts prevent aplastic anaemia or agranulocytosis, when the same boxed warning says minor haematological changes are unlikely to signal either
Why it sounds right. It follows from something real on this page: a body step, an animal result or a related finding.
The missing step. Nobody measured this in people for this use, or the study that did measure it did not show it.
What would change this
A study in people, for this exact use, measuring this exact thing, that a reviewer signs off.
RNAWiki has not yet published a reviewed conclusion for this use.
✗Goes past the evidence
That lamotrigine controls focal seizures better than carbamazepine, when SANAD found the opposite direction on the remission endpoint
Why it sounds right. It follows from something real on this page: a body step, an animal result or a related finding.
The missing step. Nobody measured this in people for this use, or the study that did measure it did not show it.
What would change this
A study in people, for this exact use, measuring this exact thing, that a reviewer signs off.
RNAWiki has not yet published a reviewed conclusion for this use.
✗Goes past the evidence
That a negative HLA-B*1502 test makes carbamazepine rash-safe, when the allele predicts SJS and TEN only and HLA-A*3101 covers a separate population and a broader set of reactions
Why it sounds right. It follows from something real on this page: a body step, an animal result or a related finding.
The missing step. Nobody measured this in people for this use, or the study that did measure it did not show it.
What would change this
A study in people, for this exact use, measuring this exact thing, that a reviewer signs off.
RNAWiki has not yet published a reviewed conclusion for this use.
What is missing or unclear◇Read from sources, not yet reviewed
What nobody knows yet
Open questions, each with why it is open and what would close it.
Missing populations
Which groups were under-represented in the studies has not been recorded.
Why it matters. Recorded by the evidence model as a gap on this record.
What would answer it
A study that measures it, and a reviewer to sign it off.
Last checked 2026-09-04. Searched: The sources listed in the receipts at the foot of this page.
Missing long-term data
No completed tested study window is recorded.
Why it matters. Recorded by the evidence model as a gap on this record.
What would answer it
A study that measures it, and a reviewer to sign it off.
Last checked 2026-09-04. Searched: The sources listed in the receipts at the foot of this page.
No reviewed conclusion
RNAWiki has not yet published a reviewed conclusion for this use.
Why it matters. Recorded by the evidence model as a gap on this record.
What would answer it
A study that measures it, and a reviewer to sign it off.
Last checked 2026-09-04. Searched: The sources listed in the receipts at the foot of this page.
Missing interaction studies
No interaction was found in the registers checked. Not finding one is not the same as showing there is none.
Why it matters. Recorded by the evidence model as a gap on this record.
What would answer it
A study that measures it, and a reviewer to sign it off.
Last checked 2026-09-04. Searched: The sources listed in the receipts at the foot of this page.
Formulation uncertainty
Several salts, forms or products of Carbamazepine are recorded. Results from one form may not transfer to another.
Why it matters. Recorded by the evidence model as a gap on this record.
What would answer it
A study that measures it, and a reviewer to sign it off.
Last checked 2026-09-04. Searched: The sources listed in the receipts at the foot of this page.
Mechanism not reviewed
No reviewed mechanism story exists for this substance.
Why it matters. Recorded by the evidence model as a gap on this record.
What would answer it
A study that measures it, and a reviewer to sign it off.
Last checked 2026-09-04. Searched: The sources listed in the receipts at the foot of this page.
What happened in people◇Read from sources, not yet reviewed
Check any of this yourself
Every line above traced back to the study it came from. This is the technical layer, and the vocabulary changes here.
Every line above can be traced to the study named beside it. Follow the link and read it.
Veterans Affairs 1985: best overall result of the four drugs then available
In plain words
Six hundred and twenty-two adults were assigned at random and double-blind to one of four anti-seizure drugs and followed for two years. Carbamazepine and phenytoin came out on top; primidone came last, mostly because people could not tolerate it.
What was measured
Overall treatment success at two years, combining seizure control and drug tolerability
Effect estimate
No reviewed effect estimate exists for this record.
Limits
See the technical detail.
From the source
Mattson and colleagues ran a 10-centre double-blind trial in 622 adults with partial and secondarily generalised tonic-clonic seizures, randomised to carbamazepine, phenobarbital, phenytoin or primidone and followed for two years or until failure. Overall treatment success was highest with carbamazepine or phenytoin, intermediate with phenobarbital, lowest with primidone (p<0.002), and the difference was driven mainly by primidone causing more nausea, vomiting, dizziness and sedation. Control of tonic-clonic seizures did not differ significantly between drugs. Carbamazepine gave complete control of partial seizures more often than primidone or phenobarbital (p<0.03). Phenytoin caused more dysmorphic effects and hypersensitivity.
Written into the record, not signed off as a reviewed claim
Veterans Affairs 1992: better than valproate for complex partial seizures
In plain words
A second blinded trial of 480 adults compared carbamazepine with valproate. For seizures that spread into a convulsion the two were equal. For complex partial seizures, four of five measures favoured carbamazepine.
What was measured
Seizure counts, time to first seizure and a composite control-plus-toxicity score over one to five years
Effect estimate
No reviewed effect estimate exists for this record.
Limits
See the technical detail.
From the source
VA Cooperative Study 264 randomised 480 adults double-blind to carbamazepine or divalproex sodium, dosed to mid-therapeutic blood levels, followed one to five years. For secondarily generalised tonic-clonic seizures the drugs were comparable (136 and 138 patients). For complex partial seizures carbamazepine was favoured on total seizure number (2.7 versus 7.6, p=0.05), seizures per month (0.9 versus 2.2, p=0.01), time to first seizure (p<0.02), seizure-rating score (p=0.04) and a composite score combining control and adverse effects (p<0.001). Valproate caused more weight gain above 5.5 kg (20% versus 8%, p<0.001), hair loss or texture change (12% versus 6%, p=0.02) and tremor (45% versus 22%, p<0.001). Rash was more common on carbamazepine (11% versus 1%, p<0.001).
Written into the record, not signed off as a reviewed claim
SANAD: lamotrigine beat it on the outcome that counts stopping the drug
In plain words
The largest first-line trial ever run in focal epilepsy put carbamazepine against four newer drugs in 1,721 patients. Lamotrigine was better on how long people stayed on their assigned drug, and that result moved carbamazepine out of first place in United Kingdom guidance.
What was measured
Time to treatment failure and time to 12-month remission, both co-primary
Effect estimate
No reviewed effect estimate exists for this record.
Limits
This entry records a failure.
From the source
SANAD arm A was an unblinded randomised trial in 1,721 patients for whom carbamazepine was deemed standard treatment, assigned to carbamazepine, gabapentin, lamotrigine, oxcarbazepine or topiramate, with co-primary outcomes of time to treatment failure and time to 12-month remission. For time to treatment failure lamotrigine was significantly better than carbamazepine (HR 0.78, 95% CI 0.63 to 0.97), gabapentin (0.65, 0.52 to 0.80) and topiramate (0.64, 0.52 to 0.79). For time to 12-month remission the estimates ran the other way and did not reach significance: carbamazepine against lamotrigine HR 0.91 (0.77 to 1.09), against topiramate 0.86 (0.72 to 1.03), against oxcarbazepine 0.92 (0.73 to 1.18); carbamazepine was significantly better than gabapentin (0.75, 0.63 to 0.90). The per-protocol difference in 12-month remission between lamotrigine and carbamazepine was 0 percentage points (95% CI -8 to 7) at two years and 5 (-3 to 12) at four.
Written into the record, not signed off as a reviewed claim
A genetic screen made Stevens-Johnson syndrome preventable, prospectively
In plain words
In Taiwan, 4,877 people about to start carbamazepine were genotyped first. The 7.7% who carried a particular immune gene variant were given something else. Among everyone else who took the drug, not one case of the life-threatening skin reaction occurred, where about ten were expected.
What was measured
Incidence of Stevens-Johnson syndrome and toxic epidermal necrolysis in HLA-B*1502-negative patients started on carbamazepine, against a historical control rate
Effect estimate
No reviewed effect estimate exists for this record.
Limits
See the technical detail.
From the source
Chen and colleagues recruited 4,877 carbamazepine-naive candidates from 23 Taiwanese hospitals and genotyped HLA-B*1502. Carriers (7.7%) were advised against carbamazepine; non-carriers (92.3%) were advised to take it and were interviewed weekly for two months. Mild transient rash developed in 4.3% and more widespread rash requiring hospitalisation in 0.1%. No SJS or TEN developed in any HLA-B*1502-negative subject receiving carbamazepine, against roughly ten cases predicted by the historical incidence of 0.23% (p<0.001). This was a historically controlled prevention study, not a randomised one, which is the correct design here only because randomising carriers to the drug would be unethical, and it is the reason the result is stated as a strong association rather than a causal effect estimate.
Written into the record, not signed off as a reviewed claim
From an unpredictable idiosyncratic reaction to a test you run before the first dose
In plain words
For forty years the severe rash on this drug was treated as bad luck that nobody could foresee. In 2004 a one-page report in Nature found a genetic marker for it in Han Chinese patients, and by 2007 the United States label required testing before treatment in people of Asian ancestry.
What was measured
That "idiosyncratic" means causeless. It meant unexplained, and in this case the explanation was findable and is now printed in the boxed warning.
Effect estimate
No reviewed effect estimate exists for this record.
Limits
See the technical detail.
From the source
Chung and colleagues reported in 2004 a strong association in Han Chinese between HLA-B*1502 and carbamazepine-induced Stevens-Johnson syndrome, and proposed that the association could be turned into a predictive test. The FDA added an HLA-B*1502 screening recommendation to the carbamazepine boxed warning in December 2007. The current label states that the allele exceeds 15% prevalence in Hong Kong, Thailand, Malaysia and parts of the Philippines, is about 10% in Taiwan and 4% in North China, 2 to 4% in South Asians, under 1% in Japan and Korea, and largely absent in people not of Asian origin. A second and separate association, HLA-A*3101, was reported in 2011 in European, Korean and Japanese ancestry and covers hypersensitivity reactions more broadly, including maculopapular eruption and DRESS, which HLA-B*1502 does not predict. The pharmacological understanding of the drug did not change. What changed was that a risk previously described as idiosyncratic acquired a population-specific, testable cause.
Written into the record, not signed off as a reviewed claim
The mechanism everyone teaches is not the mechanism the label states
In plain words
Textbooks describe carbamazepine as a use-dependent sodium channel blocker. Its own label describes reduced polysynaptic responses and blocked post-tetanic potentiation, and then says the mechanism of action remains unknown.
What was measured
That sodium-channel block is the established mechanism of carbamazepine in people, and that the parent drug is the active agent
Effect estimate
No reviewed effect estimate exists for this record.
Limits
See the technical detail.
From the source
The United States prescribing information section on mechanism of action reports anticonvulsant activity in electrically and chemically induced rodent seizures, reduction of polysynaptic responses, block of post-tetanic potentiation, abolition of pain from infraorbital nerve stimulation in cats and rats, and depression of thalamic potential and bulbar and polysynaptic reflexes. It then states in as many words that the mechanism of action remains unknown. It also notes that the principal metabolite, carbamazepine-10,11-epoxide, has anticonvulsant activity in animal models, that clinical activity for the epoxide has been postulated, and that the significance of that activity for the safety and efficacy of carbamazepine has not been established. The state-dependent sodium-channel account is well supported in cellular electrophysiology; what has never been established is the quantitative link from that block to a suppressed seizure in a person, or how much of the clinical effect belongs to the parent drug rather than the epoxide.
Source
Carbamazepine United States prescribing information, Clinical Pharmacology, Mechanism of Action (openFDA drug label endpoint)
Role in the trial
Not matched to a registered study
Identity check
no open identity hold
Audit mark
caution
Review state
Written into the record, not signed off as a reviewed claim
It doubles its own clearance in the first month, then keeps doing it to other drugs
In plain words
Carbamazepine turns on the liver enzymes that destroy it. Over three to five weeks its half-life falls from as long as 65 hours to as little as 12, so a dose that worked at the start stops working. It does the same to most other medicines a person takes.
What was measured
Elimination half-life before and after autoinduction, and the named list of interacting agents on the label
Effect estimate
No reviewed effect estimate exists for this record.
Limits
See the technical detail.
From the source
The label states that autoinduction is complete after 3 to 5 weeks of a fixed dosing regimen, with initial half-life values of 25 to 65 hours falling to 12 to 17 hours. Carbamazepine is 76% plasma protein bound with a CSF to serum ratio of 0.22. It is a potent inducer of CYP3A4 and of UGT enzymes, so it lowers the exposure of hormonal contraceptives, direct oral anticoagulants, many antiretrovirals and several other anti-seizure drugs. In the other direction the label names aprepitant, cimetidine, ciprofloxacin, danazol, diltiazem, macrolides, fluoxetine, fluvoxamine, trazodone, omeprazole, oxybutynin, isoniazid, nicotinamide, azole antifungals, acetazolamide, verapamil, ticlopidine, grapefruit juice and protease inhibitors as agents that raise carbamazepine levels, and loxapine, quetiapine, valproic acid and brivaracetam as agents that raise the epoxide by inhibiting microsomal epoxide hydrolase.
Source
Carbamazepine United States prescribing information, Clinical Pharmacology and Drug Interactions (openFDA drug label endpoint)
Role in the trial
Not matched to a registered study
Identity check
no open identity hold
Audit mark
verified
Review state
Written into the record, not signed off as a reviewed claim
Routine blood counts are monitored on a warning the label says they do not predict
In plain words
The boxed warning names aplastic anaemia and agranulocytosis, and most patients have blood counts checked because of it. The same warning says the minor changes those counts pick up are unlikely to signal either condition.
What was measured
Relative risk of 5 to 8 against a baseline of 6 and 2 cases per million per year
Effect estimate
No reviewed effect estimate exists for this record.
Limits
See the technical detail.
From the source
The boxed warning cites a population-based case-control study finding a 5 to 8 fold increased risk of aplastic anaemia and agranulocytosis on carbamazepine, against an untreated general-population rate of approximately six cases of agranulocytosis and two of aplastic anaemia per million people per year. It then states that although transient or persistent decreases in platelet or white cell counts are not uncommon, data are not available to estimate their incidence or outcome accurately, that the vast majority of leukopenia cases have not progressed, and that because the incidence of the serious conditions is so low, the vast majority of minor haematological changes seen on monitoring are unlikely to signal either abnormality. A pretreatment baseline is called for; the warning does not establish that a schedule of repeat counts detects the events it names.
Source
Carbamazepine United States prescribing information, boxed warning, Aplastic Anemia and Agranulocytosis (openFDA drug label endpoint)
Role in the trial
Not matched to a registered study
Identity check
no open identity hold
Audit mark
caution
Review state
Written into the record, not signed off as a reviewed claim
EURAP: 5.5% major malformation rate, roughly double lamotrigine and levetiracetam
In plain words
Across 1,957 pregnancies on carbamazepine alone in a 42-country registry, 5.5% of babies had a major birth defect. Lamotrigine and levetiracetam were at 2.9% and 2.8%; valproate was at 10.3%.
What was measured
Prevalence of major congenital malformations at 1 year, by drug and dose
Effect estimate
No reviewed effect estimate exists for this record.
Limits
See the technical detail.
From the source
The EURAP prospective registry followed pregnancies on anti-epileptic monotherapy at conception from 42 countries between 1999 and 2016. Major congenital malformation prevalence at one year was 107 of 1,957 (5.5%) for carbamazepine, against 17 of 599 (2.8%) for levetiracetam, 74 of 2,514 (2.9%) for lamotrigine, 10 of 333 (3.0%) for oxcarbazepine, 6 of 152 (3.9%) for topiramate, 8 of 125 (6.4%) for phenytoin, 19 of 294 (6.5%) for phenobarbital and 142 of 1,381 (10.3%) for valproate. The authors placed lamotrigine, levetiracetam and oxcarbazepine within the background range for unexposed offspring and carbamazepine outside it. This is a registry, not a randomised comparison: drug choice tracked seizure type and severity, and the endpoint is structural malformation at one year rather than cognition later.
This order is fixed in code and does not count clicks or time on the page.
What is not here
7 questions this page could not answer
These sections were prepared and left out because there was nothing to put in them. They are listed rather than hidden, so the gaps in this record are visible.
What was measured, goal by goal — found nothing in the sources checked.
How close this is to real life — found nothing in the sources checked.
Felt, measured, or meaningful — found nothing in the sources checked.
How long anything takes — found nothing in the sources checked.
What it may clash with — found nothing in the sources checked.
Other ways to the same goal — found nothing in the sources checked.
What changed on this page — found nothing in the sources checked.
The record as stored
The full record, for auditing
Everything above is built from what is below. This layer keeps the technical vocabulary, record identifiers and every stored row, so a reader who wants to check the page can.
The older medicine-wide conclusion held in this record
Written for a clinical reader, and about the medicine as a whole rather than about one indication, population and set of trials. RNAWiki does not treat it as a programme conclusion and no reviewer has signed it off in that form. It is here word for word because it is what the record holds.
A sodium-channel blocker whose own label still says the mechanism is unknown, which matched or beat every 1980s alternative for focal seizures in a 622-patient blinded trial, lost to lamotrigine on treatment failure in the 1,721-patient SANAD arm, and whose most consequential result is not a seizure count at all but a genetic screen: no case of Stevens-Johnson syndrome occurred among 4,877 Taiwanese patients when HLA-B*1502 carriers were steered away from the drug.
Recorded evidence blocks (9)
Q2
On the Carbamazepine label: indicated for what?
"Epilepsy Carbamazepine tablets are indicated for use as an anticonvulsant drug. Evidence supporting efficacy of carbamazepine as an anticonvulsant was derived from active drug-controlled studies that enrolled patients with the following seizure types: 1.": indications and usage on Carbamazepine's label. DailyMed label · bf0d012f-da14-8ce3-d74f-56172cddebc5 · 2026-08-26
Q3
104 registered trials of Carbamazepine — at which phases?
Registered studies posting no result
83 of 104
104 registered studies of Carbamazepine: 52 phase1, 24 phase4, 8 na, 8 phase3, 7 phase2, 5 na or unstated, 1 early phase1. CLINICALTRIALS_SNAPSHOT · 2026-09-01
1190 with a PubMed record
Show the evidence
phase1
52
phase4
24
na
8
phase3
8
phase2
7
na or unstated
5
7 more recorded rows
early phase1
1
completed
80
unknown
10
active not recruiting
5
terminated
5
not yet recruiting
3
recruiting
1
recorded 2026-09-01 · last checked 2026-09-04
Q4
4 of Carbamazepine's trials stopped: futility/efficacy, other?
25 to 65 hours; Initial half-life values range from 25 to 65 hours, decreasing to 12 to 17 hours on repeated doses.
bioavailabilitypharmacokinetics
89 %; The bioavailability of the extended-release tablet was 89% compared to suspension.
metabolismpharmacokinetics
Because carbamazepine induces its own metabolism, the half-life is also variable.
recorded 2026-08-26 · last checked 2026-09-04
Q7
Which 39 trials of Carbamazepine posted no result?
Posted no result
39 of 39 completed trials
Registrations
NCT00000191, NCT00000242, NCT00207428, NCT00000441, NCT00007670 and NCT00153296, and 33 more
Completion dates
oldest 1994-02; newest 2024-06-18
Show the evidence
Trial
NCT00000191
1994-02
NCT00000242
1997-01-25
NCT00207428
2000-04
NCT00000441
2000-12
NCT00007670
2003-03
NCT00153296
2005-04
14 further recorded trials
NCT00000218
2005-06
NCT00190892
2006-06
NCT00203567
2008-05
NCT00913107
2008-06
NCT00896987
2008-12
NCT01030094
2009-04
NCT00894010
2011-06
NCT01365403
2011-07
NCT01382017
2012-08
NCT01629368
2012-09
NCT01635829
2012-09
NCT01707407
2012-11-01
NCT02132897
2014-06
NCT02705768
2017-03
Q8
At the median, Carbamazepine's trials enrolled 46 people — anything larger?
Median enrolment
46
Largest enrolment
1037352
Registered trials counted
102
Q9
What do 7246 spontaneous reports say about Carbamazepine — and not say?
These are reports people sent to a regulator. They do not show the medicine caused the reaction. Nobody counted how many people took the medicine and reported nothing. The same event can be reported more than once, and many reports are incomplete. News coverage, lawsuits and new warnings change how often people report. A count is not a rate and not a risk.
Carbamazepine appears in spontaneous reports to regulators. Across the 10 most-reported reaction terms, 7246 reaction mentions were counted: pyrexia 1142; convulsion 841; drug interaction 816; toxicity to various agents 765. FAERS via Open Targets · CHEMBL108 · 2026-06-24
Show the evidence
pyrexia
1142
convulsion
841
drug interaction
816
toxicity to various agents
765
seizure
739
somnolence
679
4 more recorded rows
hyponatraemia
675
drug reaction with eosinophilia and systemic symptoms
608
epilepsy
541
balance disorder
440
recorded 2026-06-24 · last checked 2026-09-04
Q10
Which 10 reactions does Carbamazepine's label not list?
ChEMBL 37 — CC BY-SA 3.0 Unported · ChEMBL ATC CC BY-SA · ClinicalTrials.gov — US Government work · Open Targets 26.06 — CC0 · mixed register set; per-register licences in docs/specs/corpus-20k-sources.md · openFDA / DailyMed — US Government work
✗ required summary fields resolved: 4 required field(s) not terminal: Why people use it, Best-supported result, Biggest unanswered question, Human evidence
✓ public claims reviewed: 0 reviewed claim(s); drafts are never rendered
✓ source coverage passed: 7 source rows
✓ no critical contamination: no quarantine open
✓ canonical metadata passed: slug and display name present
✓ no raw internal fields: enforced by the copy-contract test over the rendered page
This is a record of evidence. It is not medical advice, and it does not say this substance suits you. Nothing here says any substance on RNAWiki is appropriate for a child.