# Optimizing antiseizure medication deprescribing decisions across the lifespan by integrating seizure risk prediction and patient preferences towards individualized net-benefit assessment

> **NIH NIH K23** · UNIVERSITY OF MICHIGAN AT ANN ARBOR · 2024 · $196,395

## Abstract

PROJECT SUMMARY / ABSTRACT
 Objectives. A critical need remains for comprehensive, personalized strategies for considering
deprescribing across the lifespan. My long-term goal is to improve pharmacologic stewardship of neurologic
medications to promote healthy aging using decision science techniques. This application’s objective is to
collect preliminary data for a future large, multicenter RCT that will capture how the absolute effect of
discontinuation changes across key subgroups such as evolving healthcare preferences across life (SA1) and
different levels of seizure risk (SA2).
 Aims/Methods. First, we will define how patient preferences influencing ASM decisions change across
the lifespan. We will survey ~300 patients with epilepsy at 4 sites, with stratified sampling across ages. The
survey will include best-worst scaling ranking and time trade-off rating exercises, plus more general questions
assessing willingness to enroll in our future deprescribing trial. A deeper understanding of how patients rank
the importance of seizure relapse versus other factors would more closely align clinicians with patient values,
and therefore enhance our ability to deliver patient-centered values-concordant care. Second, we will improve
post-deprescribing seizure prediction. We will update the prior individualized post-discontinuation seizure risk
calculator by adding modern adult data and statistical techniques for improved out-of-sample predictions in
3,147 international patients. We will then execute decision analytic models to assess which patients might
experience theoretical net benefit versus harm from deprescribing ASMs and the effects of different population-
wide deprescribing strategies. This modeling would improve clinicians’ ability to deliver personalized medicine,
plus facilitate our future trial by developing analytic procedures (the future trial will assess how low seizure risk
must be until deprescribing exerts predicted benefit on quality of life) and justifying which patients to enroll.
Third, we will refine our trial protocol and test procedures at 3 sites for our future planned trial in which we will
randomize seizure-free patients to deprescribing ASMs. This work will result in more personalized evidence-
based recommendations to tailor ASM deprescribing decisions to a person’s risk, healthcare preferences, and
stage of life. This will lead to many R01’s including but not limited to applying these techniques to other
deprescribing decisions particularly as they pertain to neurological medications used by older patients.
 Candidate. Through expert mentorship, advanced coursework, and the intellectually rich environment
of the University of Michigan, this award will develop Dr Terman into a leader using decision sciences
synthesizing patient preferences with risk prediction, and leading clinical trials, aimed at rational
(de)prescribing of central nervous system agents tailored to how preferences and risk change across life.

## Key facts

- **NIH application ID:** 10805195
- **Project number:** 1K23AG081463-01A1
- **Recipient organization:** UNIVERSITY OF MICHIGAN AT ANN ARBOR
- **Principal Investigator:** Samuel W Terman
- **Activity code:** K23 (R01, R21, SBIR, etc.)
- **Funding institute:** NIH
- **Fiscal year:** 2024
- **Award amount:** $196,395
- **Award type:** 1
- **Project period:** 2024-08-15 → 2029-04-30

## Primary source

NIH RePORTER: https://reporter.nih.gov/project-details/10805195

## Citation

> US National Institutes of Health, RePORTER application 10805195, Optimizing antiseizure medication deprescribing decisions across the lifespan by integrating seizure risk prediction and patient preferences towards individualized net-benefit assessment (1K23AG081463-01A1). Retrieved via AI Analytics 2026-05-28 from https://api.ai-analytics.org/grant/nih/10805195. Licensed CC0.

---

*[NIH grants dataset](/datasets/nih-grants) · CC0 1.0*