Charted times of bedside pain assessments are slot labels: prevalence across hospitals, consequences for estimating analgesic onset, and when hourly charting is an adequate approximation

Objective. Temporal and causal analyses of electronic health records (EHRs) treat the charted time of an observation as the time it was made. We quantified how often bedside pain and sedation assessments are charted to a time slot, how this affects estimates of analgesic onset, and when hourly charting is an adequate approximation. Methods. We used intensive care unit (ICU) data from one academic center (MIMIC-IV) and 78 hospitals (eICU-CRD). Slot labeling was identified from charted minutes and from the delay between charted and entry times. In MIMIC-IV, opioid administration times served as anchors, and faster onset of intravenous (IV) than oral opioids as a pharmacologic benchmark. We compared onset estimated from charted times with onset from exactly timed records, bounded true assessment times under alternative charting mechanisms, and used simulation to relate bias to the ratio of slot width to response time constant (Δ/k). Results. In MIMIC-IV, 68% of pain and 76% of sedation scores were charted on the hour and entered a median of 41–44 minutes later. Of the first reassessments after a dose, 32–34% of on-hour records carried a time before the dose, against 12–14% of off-hour records. Across eICU hospitals, the on-hour share ranged from 15% to 92% (10th–90th percentile). By charted time, oral opioids appeared nearly as fast as IV (pain change at 15–30 minutes, −1.76 vs −2.11 points); exactly timed records showed −0.80 vs −2.00. Slower oral onset held under every charting mechanism compatible with the data, but its magnitude was not identified (ratio of time constants 1.2–11.6). In simulation, analysis by charted time was adequate up to Δ/k ≈ 2 when rounding was to the nearest slot, but only up to Δ/k ≈ 0.1 when rounding had a direction. Conclusion. Charted times of bedside assessments are often slot labels, and charting practice is hospital-specific. For ICU analgesic onset, neither condition for an adequate approximation can be assumed, and EHR data alone cannot recover the size of the effect.

Authors

Institutions

Publication Details

Journal
medRxiv
Published
2026-11-17
DOI
https://doi.org/10.64898/2026.11.14.26368116
Primary Topic
Electronic Health Records Systems
Type
preprint
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
preprint

Charted times of bedside pain assessments are slot labels: prevalence across hospitals, consequences for estimating analgesic onset, and when hourly charting is an adequate approximation

Kris S. Moe, Yangming Lee
1 links
medRxiv
Electronic Health Records Systems
preprint

Charted times of bedside pain assessments are slot labels: prevalence across hospitals, consequences for estimating analgesic onset, and when hourly charting is an adequate approximation

Kris S. Moe, Yangming Lee
preprint en

Abstract

Objective. Temporal and causal analyses of electronic health records (EHRs) treat the charted time of an observation as the time it was made. We quantified how often bedside pain and sedation assessments are charted to a time slot, how this affects estimates of analgesic onset, and when hourly charting is an adequate approximation. Methods. We used intensive care unit (ICU) data from one academic center (MIMIC-IV) and 78 hospitals (eICU-CRD). Slot labeling was identified from charted minutes and from the delay between charted and entry times. In MIMIC-IV, opioid administration times served as anchors, and faster onset of intravenous (IV) than oral opioids as a pharmacologic benchmark. We compared onset estimated from charted times with onset from exactly timed records, bounded true assessment times under alternative charting mechanisms, and used simulation to relate bias to the ratio of slot width to response time constant (Δ/k). Results. In MIMIC-IV, 68% of pain and 76% of sedation scores were charted on the hour and entered a median of 41–44 minutes later. Of the first reassessments after a dose, 32–34% of on-hour records carried a time before the dose, against 12–14% of off-hour records. Across eICU hospitals, the on-hour share ranged from 15% to 92% (10th–90th percentile). By charted time, oral opioids appeared nearly as fast as IV (pain change at 15–30 minutes, −1.76 vs −2.11 points); exactly timed records showed −0.80 vs −2.00. Slower oral onset held under every charting mechanism compatible with the data, but its magnitude was not identified (ratio of time constants 1.2–11.6). In simulation, analysis by charted time was adequate up to Δ/k ≈ 2 when rounding was to the nearest slot, but only up to Δ/k ≈ 0.1 when rounding had a direction. Conclusion. Charted times of bedside assessments are often slot labels, and charting practice is hospital-specific. For ICU analgesic onset, neither condition for an adequate approximation can be assumed, and EHR data alone cannot recover the size of the effect.

medRxiv
Rochester Institute of Technology (US), University of Washington (US)
Good health and well-being
Electronic Health Records Systems
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon