Antiseizure Medication Administration Gaps Across the ICU-to-Floor Transfer: A Matched Within-Patient Comparison

Objective

Whether a scheduled antiseizure medication (ASM) continues on schedule across the ICU-to-floor transfer has not been characterized. We quantified ASM administration-gap frequency across this transfer and compared it with gap frequency during matched non-transfer intervals in the same patient and drug.

Methods

In this retrospective MIMIC-IV (version 3.1) cohort study, we identified epilepsy and status-epilepticus admissions with an ICU stay followed by floor transfer and a scheduled ASM order active at ICU departure. A gap was defined as an interval exceeding 1.5 times the expected dosing interval between the last ICU dose and first floor dose, or no further dose before discharge, and compared with a matched non-transfer control interval in the same patient and drug (paired McNemar test). A multivariable model evaluated six prespecified clinical predictors; sociodemographic variables were summarized descriptively.

Results

Among 2,469 ASM transition-by-drug observations (1,583 admissions, 1,335 patients), an administration gap occurred in 251 (10.2%; 95% CI, 8.7%-11.7%). Gap frequency across the transfer exceeded frequency during matched non-transfer control intervals in the same patient and drug: a paired rate difference of 5.8 percentage points (95% CI, 4.4-7.1; 7.5% vs 1.7%; P = 7.3 × 10 ^-22 ) before the transfer and 6.4 percentage points (95% CI, 4.9-7.9; 8.9% vs 2.5%; P = 1.9 × 10 ^-23 ) after. Gap rates were similar for intravenous-available (9.9%) and oral-only (11.4%) drugs (rate difference, 1.5 percentage points; 95% CI, -1.6 to 4.5; P = .34). None of six prespecified predictors reached significance after correction.

Significance

An antiseizure medication administration gap occurred in approximately 1 of every 10 drug-transition observations at the ICU-to-floor transfer, exceeding matched non-transfer gap rates by 5.8 to 6.4 percentage points. This transfer-associated excess, rather than any single medication or patient characteristic, supports a structured medication-continuity check.

Key Points

• Antiseizure medication administration gaps occurred in 10.2% of ICU-to-floor drug transitions (95% CI, 8.7%-11.7%).

• Gap frequency was 5.8-6.4 percentage points higher across the transfer than during matched non-transfer control intervals.

• No prespecified clinical predictor or IV-availability class identified a clearly higher-risk subgroup.

• Gap rates were similar for intravenous-available and oral-only medications (9.9% vs 11.4%; difference, 1.5 points; 95% CI, -1.6 to 4.5).

• A structured medication-continuity check at the ICU-to-floor transfer may reduce this administration gap.