Drowsy Driving in Fatal Crashes, United States, 2017–2021
archive: archived pipeline: cataloged verified
Get this paper ↗ (full text — opens at the source; we link to it, we don't host it)
Summary
This study addresses the significant underestimation of drowsy driving’s role in motor vehicle fatalities, a problem arising from police reports that often fail to capture driver alertness. While the National Highway Traffic Safety Administration (NHTSA) acknowledges that police-reported data substantially undercount drowsy driving incidents, previous estimates lacked the temporal and spatial resolution needed to track trends or evaluate countermeasures. The AAA Foundation for Traffic Safety conducted this research to provide updated national estimates for 2017–2021 and to develop a methodology capable of analyzing yearly changes and regional variations. To achieve this, the researchers developed and validated a multiple imputation model using data from the NHTSA’s Crash Investigation Sampling System (CISS), which contains in-depth investigations of crashes involving towed passenger vehicles. In CISS, investigators often could not determine driver drowsiness, particularly in fatal cases. The model used logistic regression to impute drowsiness status based on variables such as driver age, sex, time of day, crash type, and alcohol use. This model was validated by simulating unknown drowsiness cases within the CISS dataset, showing high accuracy. The validated model was then applied to the Fatality Analysis Reporting System (FARS) to impute drowsiness for all passenger vehicle drivers involved in fatal crashes nationwide, treating all FARS drowsiness data as unknown to ensure consistency. The results indicate that an estimated 17.6% of all fatal crashes in the United States between 2017 and 2021 involved a drowsy driver, resulting in 29,834 fatalities. This figure is more than seven times higher than estimates derived solely from police reports. The percentage of fatal crashes involving drowsy driving remained relatively constant over the five-year period, but the absolute number of fatalities increased significantly due to a rise in total fatal crashes. Drowsiness was most prevalent among drivers aged 16–20, though the largest number of drowsy drivers were aged 21–34. Men were significantly more likely to be drowsy than women. Notably, approximately one-third of drowsy drivers in fatal crashes had non-zero blood alcohol concentrations, and drivers with alcohol in their system were nearly twice as likely to be drowsy as sober drivers. However, two-thirds of drowsy drivers had not been drinking. Head-on crashes and road departures were the most common crash types associated with drowsy driving. The study concludes that drowsy driving remains a major contributor to traffic fatalities, with current official statistics severely underreporting its prevalence. The developed imputation methodology offers a robust tool for future research, enabling the examination of trends over time and the evaluation of countermeasures with greater precision than previously possible. The findings underscore the need for more effective strategies to combat drowsy driving, particularly given the significant overlap between drowsiness and alcohol impairment, and the high proportion of drowsy drivers who were not under the influence of alcohol.
Key finding
An estimated 17.6% of all fatal motor vehicle crashes in the United States between 2017 and 2021 involved a drowsy driver, resulting in approximately 29,834 fatalities.
Methodology
dataset
Sample size: 208727
Provenance
The full processing record for this entry. Every stage of this paper's journey through the pipeline is logged — what ran, with which tool and model, how many attempts it took, and when it last completed. Discovered via bulk_ingest_aaa_foundation on 2026-05-23 (6 acquisition events logged).
| Stage | Outcome | Tool | Model | Prompt | Attempts | Completed |
|---|---|---|---|---|---|---|
| discover | success | aaa_foundation | — | — | 2 | 2026-05-23 |
| archive | success | — | — | — | 1 | 2026-05-23 |
| extract | success | cached | — | — | 2 | 2026-06-10 |
| clean | success | — | — | — | 1 | 2026-06-01 |
| chunk | success | — | — | — | 1 | 2026-06-01 |
| embed | success | — | — | — | 1 | 2026-06-02 |
| enrich | success | — | — | — | 1 | 2026-05-23 |
| promote | success | — | — | — | 1 | 2026-05-23 |
| summarize | success | llm | qwen3.6-27b-prismaquant | summ-v5 | 3 | 2026-06-10 |
| tag | success | vector_similarity | — | — | 19 | 2026-06-11 |
| verify | success | — | — | — | 2 | 2026-06-10 |
Summary generated by qwen3.6-27b-prismaquant on 2026-06-10; verification: verified.
Topics
Ranked by relevance to this paper. Hover a topic for its definition.
- drowsy as impairment
- drowsiness detection algorithms
- drowsiness
- sleep deprivation
- incidence prevalence
- truck driver fatigue
Information type
What kind of knowledge this paper contributes, grouped by family — independent of topic (what it is about) and method (how it was studied).
- Empirical Findings: crash risk outcomes, observational prevalence
- Methodological Resource: dataset resource