Audio Epilepsy Digest Episode 018 · Source Review

AI-Assisted Editorial Review

Episode 18 Source Review

The review checked a long-term medication-safety discussion against twelve full-text sources, corrected causal, denominator, and comparative overstatements across four withheld drafts, and retained focused caveats for clinician listening.

This is not independent peer review. It is an internal editorial check — AI-assisted — to verify that the spoken audio stays faithful to what the published studies actually found.

Review Goal

The review asked a narrow question:

Does the spoken audio stay faithful to the source papers, especially where clinical interpretation could be overstated?

The review focused on the difference between years of clinical experience and years of structured observation, common versus rare effects, survivor-enriched continuation cohorts, comparative tolerability, front-loaded lamotrigine rash risk, lacosamide conduction susceptibility, and the boundary between absence of a cumulative-toxicity signal and proof of lifetime safety.

Evidence Standard

The episode was reviewed against the full text of the source articles — not abstracts or summaries — to catch places where the audio could misrepresent scope, effect size, or clinical implications.

  1. Current KEPPRA prescribing information. DailyMed (updated 2026).
  2. Current LAMICTAL prescribing information. DailyMed (current version 2025).
  3. Current VIMPAT prescribing information. DailyMed (updated 2026).
  4. Marson AG, et al. "The SANAD II study of the effectiveness and cost-effectiveness of levetiracetam, zonisamide, or lamotrigine for newly diagnosed focal epilepsy." The Lancet (2021). PMID: 33838757.
  5. Fröling M, et al. Systematic review and meta-analysis of adverse events with levetiracetam and brivaracetam. (2026). PMID: 41824075.
  6. Depondt C, et al. "The long term retention of levetiracetam in a large cohort of patients with epilepsy." Journal of Neurology, Neurosurgery & Psychiatry (2006). PMID: 16361605.
  7. Bauer J, et al. Long-term levetiracetam follow-up in selected prior beneficiaries. (2006). PMID: 16911344.
  8. Faught E, et al. Six-year continuation experience with lamotrigine. (2004). PMID: 14751204.
  9. Li J, et al. Real-world comparison of oxcarbazepine, lamotrigine, and levetiracetam in newly diagnosed focal epilepsy. (2020). PMID: 32679487.
  10. Ben-Menachem E, et al. Long-term lacosamide versus controlled-release carbamazepine extension study. Epilepsia (2019). PMID: 31755090.
  11. Rosenfeld W, et al. Long-term open-label lacosamide extension up to eight years. (2014). PMID: 25461210.
  12. Buoso C, et al. Review and meta-analysis of antiseizure medications and bone health. Acta Endocrinologica (2024). PMID: 40530090.

Process

Step 1Generate from full text

The episode was generated directly from the source papers, with explicit instructions to preserve study scope and avoid overstating findings.

Step 2Transcribe audio

The audio was converted to text so spoken claims could be checked line by line against the papers.

Step 3Audit claims twice

Two independent AI-assisted reviews checked spoken claims against the source text and flagged material overstatement, misrepresentation, or unsupported interpretation.

Each complete draft received two independent reviews and one combined editorial verdict: cleared, cleared with minor notes, needs revision, or re-record required. A draft could not advance until accuracy issues were resolved.

Verdict History

DraftVerdictWhat changed
First draftCleared with minor notesThe original shorter episode cleared review, then the editorial scope was reopened to add explicit approval-history and longest-observed-horizon evidence requested for patient counseling.
Second draftNeeds revisionIt treated a proposed SV2A explanation as established causation, implied that late serious lamotrigine rash was impossible, recommended universal bone monitoring, and ended with unsupported future prediction.
Third draftRe-record requiredIt corrected those issues but reversed the direction of the Li paper's own conclusion about which drug it called safest.
Fourth draftRe-record requiredIt drifted away from the patient's Keppra question into a general bone-health episode and converted heterogeneous evidence into individual causal risk and treatment advice.
Fifth draftCleared with minor notesIt restored the direct patient question, the two evidence clocks, all three medication profiles, the study denominators, and a bounded answer without a lifetime guarantee or universal ranking.

How the Language Changed

These are the specific spoken claims that failed review, shown alongside the corrected language in the published draft. Quoted or closely paraphrased from the transcripts.

Mechanism and timing Fixed: Draft 2 → 3

Draft 2 said

SV2A binding caused levetiracetam irritability, and reaching prolonged lamotrigine exposure made serious rash impossible.

Published version says

The behavioral mechanism is proposed rather than established; lamotrigine serious-rash risk is front-loaded, but later risk is not literally zero.

Comparative interpretation Fixed: Draft 3 → 4

Draft 3 said

The Li paper crowned lamotrigine the safest option.

Published version says

The paper called levetiracetam safest, while the episode explains why its nonrandom selection and nonsignificant adjusted comparison do not support a universal ranking.

Patient question and cumulative risk Fixed: Draft 4 → 5

Draft 4 said

Heterogeneous bone-health findings justified an individual fracture estimate, universal supplementation, and a broad systemic-harm conclusion.

Published version says

Bone health is a bounded example of what may accumulate, while the direct Keppra answer returns to behavior, fatigue, renal exposure, rare reactions, seizure benefit, and continued reassessment.

Final Result

Verdict: Cleared with minor notes

Both independent reviews found no unresolved meaning-changing source-fidelity or clinical-safety defect in the fifth draft; the clinician-editor then completed listening QA and authorized release.

Minor caveats carried into the show notes

  • The proposed relationship between SV2A biology and irritability remains a hypothesis, not an established causal mechanism.
  • The longest follow-up figures are maximum horizons in selected continuation cohorts, not the typical exposure for every participant or a lifetime safety guarantee.
  • The Faught lamotrigine cohort and the Bauer and Rosenfeld extension cohorts are enriched for people who had already tolerated or benefited from treatment.
  • The bone-health evidence is a cumulative-toxicity counterexample and should not be converted into a drug-specific fracture probability for levetiracetam, lamotrigine, or lacosamide.

What Changed

  • Separated years since approval from years of structured observation.
  • Changed causal language about irritability into a tentative mechanistic hypothesis.
  • Restored the possibility of rare later lamotrigine rash while preserving its front-loaded timing.
  • Corrected the direction of the Li paper's stated conclusion and kept its nonrandomized limitations adjacent.
  • Removed universal bone-testing and supplementation instructions unsupported by the source packet.
  • Restored the patient's exact Keppra question and answered it directly before the listener sign-off.

Boundaries

This review does not certify clinical recommendations, replace human editorial judgment, replace independent expert review, transfer responsibility to the authors of the source studies, or make the episode a clinical guideline.