Cosmos Study

Hazardous Holidays? Injury Patterns in Urgent and Emergency Care Vary by Injury Type and Time of Year

October 29, 2026

Team A:Kersten Bartelt, RNJoe Deckert, PhD

Team B:Blaine Franklin, PT, DPTEric Barkley

Key Findings

Among the U.S. holidays studied, Independence Day had the largest overall increase in injuries, averaging 5,418 more injuries than baseline over the three-day holiday window. Fireworks/blast injuries rose to 22 times baseline (1,187 more cases), burns were 1.8 times baseline (1,267 more cases), and fire/smoke exposures were 2.3 times baseline (383 more cases).
Burns, eye injuries, and firework/blast injuries peak near Independence Day, while lacerations cluster around Thanksgiving and Christmas, holidays that often involve preparing meals.
Halloween 2026 falls on a Saturday. In 2025, when Halloween fell on a Friday, alcohol-related and assault visits peaked the following Saturday, well above a typical weekend. Halloween’s effect is modest in most years, but it adds to the weekend surge when the two coincide.

Holidays reshape everyday activity in ways that can change the pattern of traumatic injuries seen in emergency departments (EDs) and urgent care clinics. Injuries related to fireworks, for example, are heavily concentrated around Independence Day: previous estimates suggest roughly two-thirds of the year’s fireworks-related injuries occur in the weeks surrounding July 4.¹ Drownings also peak in summer due to recreational water exposure.² Other holidays can carry their own hazards like kitchen burns and lacerations from holiday cooking, alcohol-related injuries, or falls, but existing surveillance examines a single injury type or holiday at a time. A broader view across multiple injury types and holidays can help health systems anticipate which rates of injuries increase or decrease around specific dates.

We examined more than 56 million emergency department and urgent care injury encounters from 2022 to 2025 among U.S. patients and looked for diagnoses including burns, lacerations, fractures, firearm injuries, fireworks/blast injuries, falls, motor-vehicle injuries, drowning, and acute alcohol intoxication.

To better understand whether different injury volumes were associated with certain U.S. holidays, we counted injuries in a three-day window around the holiday (the day before, the day of, and the day after) and compared the rate with a matched window of the same length seven days removed from the holiday, established on the same days of the week and avoiding overlap with neighboring holidays. We then took the difference between the count of injuries during the holiday window and the reference window for each year between 2022 and 2025 and calculated an average change from baseline. Figure 1 illustrates an example calculation.

Figure 1

Study Design: Holiday Injury Windows and Matched Comparison Periods

Figure 1. Holiday exposure windows (±1 day) and the matched non-holiday comparison window seven days removed, used to estimate excess injuries.

Independence Day had the greatest increase in injuries, with an average of 5,418 excess injuries compared to baseline. Many of these were firework-related (1,187 additional cases, about 22 times the baseline rate), accompanied by 1,267 additional burns (about 1.8 times baseline) and 383 additional fire/smoke exposures (about 2.3 times baseline). Labor Day and New Year’s Day followed, with an average of 3,343 and 2,762 excess injuries. The three days around January 1 saw an average of 2,167 additional acute-alcohol visits (about 1.5 times baseline), 1,133 additional assaults (about 1.4 times baseline), and 1,029 additional lacerations (about 1.2 times baseline). Thanksgiving and Christmas, by contrast, saw a decrease in overall injuries from the baseline; both holidays had fewer falls, motor-vehicle injuries (MVIs), and traumatic brain injuries (TBIs) but an increase in lacerations compared to baseline.

Figure 2

Change in Emergency and Urgent Care Injuries Around Holidays by Injury Type

Figure 2. Average annual change in cases, the number of injuries during the three-day holiday window compared to the non-holiday baseline by holiday.

Daily counts show how each injury type fluctuates throughout the year. In 2025, injuries related to fireworks sat near zero for most of the year and appeared almost entirely as sharp holiday spikes, with 722 encounters on July 4 and 225 on New Year’s Day. Burns, smoke inhalation, traumatic amputation, and eye injuries also peaked around Independence Day, as shown in Figure 3.

Acute-alcohol and assault visits, by contrast, followed a weekly rhythm: alcohol-related visits ran about a third higher on weekends than on weekdays and assaults about 18% higher. As a result, a holiday that falls on or next to a weekend starts from an already elevated baseline, and its effect stacks on top of the usual weekend surge. Halloween 2025 shows this. It fell on a Friday, and the largest counts came the next day, Saturday, November 1, when alcohol (3,436) and assault (1,955) visits were well above even a typical Saturday. Averaged over 2022 to 2025 and compared with the same days of the week, Halloween’s increase was much smaller. For emergency departments and urgent care clinics, the practical takeaway is that the heaviest alcohol- and assault-related volumes are most likely when a holiday coincides with a weekend.

Figure 3

Daily Injury Volume in 2025 by Type

Figure 3. The daily injury volume across 2025 by injury type.

These data come from Cosmos, a dataset created in collaboration with a community of Epic health systems representing more than 300 million patient records from 2,200 hospitals and more than 50,000 clinics from all 50 U.S. states, Canada, Lebanon, and Saudi Arabia. This study was completed by two teams that worked independently, each composed of a clinician and research scientist. The two teams came to similar conclusions. Graphics by Brian Olson.

References

2023 Fireworks Annual Report. U.S. Consumer Product Safety Commission; 2024. https://www.cpsc.gov/s3fs-public/2023-Fireworks-Annual-Report.pdf. Accessed August 14, 2026.
Clemens T, Moreland B, Mack KA, Thomas K, Bergen G, Lee R. Vital Signs: Drowning Death Rates, Self-Reported Swimming Skill, Swimming Lesson Participation, and Recreational Water Exposure – United States, 2019-2023. MMWR Morb Mortal Wkly Rep. 2024;73(20):467-473. doi:10.15585/mmwr.mm7320e1

Data Definitions

Study period: 12/31/2021 to 12/30/2025

Study population: inclusion: An emergency or urgent care encounter within the study period for patients with a current U.S. address

Comparison baseline: For each holiday three-day window, the in-window count is compared against a window of the same length seven days removed from the holiday period. The comparison window was seven days after for all holidays except for Christmas, which was compared to seven days prior to avoid contamination from the New Year’s holiday. Shifting by seven days aligns the day-of-week effect seen in many of these injuries.

Exposures: holidays: New Year’s, Halloween, Independence Day, Thanksgiving, Christmas, St. Patrick’s Day, Mother’s Day, Memorial Day, Father’s Day, Labor Day

Exposure windows: The day prior to the holiday through the day after the holiday

Fireworks/blast injury: An encounter or billing diagnosis with ICD-10-CM code W39*, paired with a burn, eye injury, hand injury, or laceration diagnosis in the same encounter.

Burn: An encounter or billing diagnosis with ICD-10-CM code T20*–T32*

Laceration: An encounter or billing diagnosis with ICD-10-CM code S41*, S51*, S61*

Traumatic amputation (digits): An encounter or billing diagnosis with ICD-10-CM code S68*

Eye injury: An encounter or billing diagnosis with ICD-10-CM code S05*

Hand injury: An encounter or billing diagnosis with ICD-10-CM code S60*-S69* with a 7th digit of “A”

Foreign body in throat: An encounter or billing diagnosis with ICD-10-CM code T17.2*

Fracture: An encounter or billing diagnosis with ICD-10-CM code S[0-9]2* or M80*, or SNOMED code 125605004

Firearm injury: An encounter or billing diagnosis with ICD-10-CM code W32*, W33*, W34*, X72*, X73*, X74*, X93*, X94*, X95*, Y22*, Y23*, Y24*, Y35.0*, or Y38.4*, all with a 7th digit of “A” to convey an initial encounter. Alternatively, an encounter or billing diagnosis with SNOMED CT code of 219257002, 219338008, or 219334005

Traumatic brain injury: An encounter or billing diagnosis with ICD-10-CM code S06* or SNOMED code 127295002

Pedestrian injury: An encounter or billing diagnosis with ICD-10-CM code V00*-V09*

Motor-vehicle occupant injury: An encounter or billing diagnosis with ICD-10-CM code V30*-V79*

Falls: An encounter or billing diagnosis with ICD-10-CM code W00*-W19*

Assault: An encounter or billing diagnosis with ICD-10-CM code X92–X99, Y00*-Y09*

Acute alcohol intoxication: An encounter or billing diagnosis with ICD-10-CM code F10.12*, F10.22*, F10.92*, T51.0X*

Submersion/drowning: An encounter or billing diagnosis with ICD-10-CM code W65*, W67*, W69*, W73*, W74*

Fire/smoke exposure: An encounter or billing diagnosis with ICD-10-CM code X00*-X08*

Overall: Any of the studied categories. Each encounter is counted once in overall even if it falls into more than one injury category, so changes in individual categories do not sum to the overall change.

Other injury: An acute injury diagnosis (ICD-10-CM codes S00*-T88* with an initial encounter (“A”) 7th character) documented as an encounter or billing diagnosis excluding any included in other specific injury rows.

Confounder: day of week: Handled by design through the matched-weekday ±7-day comparison window (see Comparison baseline), which also accounts for holidays falling on different weekdays across years.

Stratification: age: Age groups: 0–4, 5–12, 13–17, 18–34, 35–49, 50–64, 65–79, 80+

Model specifications: Excess incidents (in-window minus baseline daily mean, × exposure days per year)
Daily counts

Limitations: External-cause codes (fireworks W39, pedestrian, drowning) are inconsistently recorded, so those categories might be under-counted.
The counts do not adjust for holiday-driven shifts in total ED/UC volume.
ED/UC only: misses direct trauma admissions and on-scene fatalities, so this reflects ED and UC injury burden, not lethality.