Abstract
Study Design
National database study.
Objective
Road traffic accidents (RTAs) are a common and challenging cause of facial fractures in the United States. The present study sought to utilize the Healthcare Cost and Utilization Project National Inpatient Sample (HCUP-NIS) to investigate national trends, injury patterns and disparities in facial fractures secondary to RTAs. To date, this is the first study to do so.
Methods
A retrospective analysis was conducted of patients with primary facial fractures secondary to RTAs using the 2018-2021 HCUP-NIS. Patients were classified into the RTA and non-RTA group. Demographics, injury patterns, and inpatient outcomes were compared.
Results
In total, 154,185 primary facial fractures were identified, of which 17% (n = 26,115) were associated with RTAs. RTAs commonly involved cars (41%), followed by motorcycles (15%), pedestrians (11%), and bicyclists (10%). The RTA group was younger (34% vs 24% < 25 years, P < .01) and more frequently Hispanic (18% vs 15%, P < .01). The most common fracture types were mandibular (23%), frontal (14%), and orbital fractures (14%). The RTA group was 50% more likely to have multiple facial fractures (OR = 1.5, P < .01). The RTA group had a longer length of stay (5.3 vs 4.0 days, P < .01), admission charge ($127,932 vs $79,414, P < .01), and mortality rate (1.9% vs 1.4%, P < .01) than the non-RTA group.
Conclusions
The present findings provide valuable insights, informing early involvement of craniofacial surgeons for the assessment of combination facial fractures and tailored treatment approaches for RTA patients.
Keywords: facial fracture, road traffic accident, motor vehicle collision, injury pattern, national database
Introduction
Road traffic accidents (RTAs) represent a common and challenging cause of facial fractures for craniofacial surgeons as they often involve high speed mechanisms and associated injuries. Studies have reported varying rates of RTA-related facial fractures, ranging from 11% to as high as 55% of all facial fractures in pediatric patients. 1 In the early 2000s, advances in vehicle safety and traffic regulations like seatbelt laws helped decrease the incidence of traffic accidents overall.2-6 However, recent reports suggest a resurgence in traffic accident injuries and fatalities due to the exacerbation of high speed, intoxicated, distracted driving, walking with increased smartphone usage, and the aftermath of the COVID-19 pandemic.7,8 Despite efforts to reduce these risks, RTAs remain the second leading cause of unintentional injury deaths in the United States, with higher rates of fatal crashes compared to other developed countries.9-11 Consequently, RTA-related facial fractures remain a pressing concern, underscoring the importance of understanding the current landscape of such injuries to inform improved treatment strategies.
Existing literature on this subject largely involves large institutional cohorts outside the United States with differing traffic landscapes 12 or national data prior to the widespread adoption of smartphones.4,13 Moreover, these studies may not encompass hospital characteristics, inpatient safety, and cost measures. As such, the aim of the present study was to utilize a large national inpatient database to provide an updated comprehensive assessment of national trends, injury patterns, and disparities of traffic-related facial fractures in the United States.
Patients and Methods
Inclusion/Exclusion Criteria
Following review by the Institutional Review Board, this study was given exempt status. A database retrospective study was conducted using the 2018, 2019, 2020, and 2021 Healthcare Cost and Utilization Project National Inpatient Sample (HCUP-NIS): the largest publicly available all payer inpatient healthcare database in the US. 14 This database includes data from more than 7 million hospital inpatient records and cover more than 97% of the US population. Patients who fulfilled the primary International Classification of Diseases, 10th Revision, Clinical Modification (ICD-10-CM) diagnosis codes for frontal/cranial, basilar skull, nasal, orbital, malar, maxillary, zygomatic, LeFort I, LeFort II, LeFort III, alveolus/tooth, mandibular, and unspecified fractures were included (Appendix I). The cohort was then stratified by the independent variable of road traffic accident (RTA) as specified by a non-primary ICD-10-CM code (Appendix II). The RTA types consisted of pedestrians, two-wheel non-motorized vehicles (“bicycles”), two-wheeled motorized vehicles (“motorcycles”), and four-wheeled motor vehicles (“cars”). All other traffic-related ICD-10-CM codes were categorized as “Other.” The cohort was further stratified by associated cervical spine injuries, including cervical spinal fractures and cervical nerve injuries, as identified by non-primary ICD-10-CM codes (Appendix III). Patients with codes for subsequent encounters or sequelae from previous facial fractures were excluded.
Data Collection
Cohort characteristics were reported for the following HCUP-NIS variables: sex, age group, race, insurance status, discharge status, residence population density, and median household income by zip code. For hospital characteristics, frequencies were calculated for hospital bed size, location/teaching status, control, and census division. Outcomes measures such as mortality rate, total admission charges, and hospital length of stay were recorded. Each patient was determined to have an isolated or combination (>1 simultaneous fracture) by determining the number of non-primary ICD-10-CM codes they had corresponding to facial fractures. The cumulative number of facial fractures was determined by the number of ICD-10 codes reporting facial fractures associated with each patient.
Statistical Analysis
Cohort characteristics, hospital characteristics, and facial fracture injury pattern variables were compared between the RTA and non-RTA group using Chi-Square or Fisher’s Exact tests. Continuous variables such as number of fractures, length of stay, and mean admission charges were compared between the RTA and non-RTA groups with independent samples t test and among transportation types with analysis of variance. In addition to the absolute incidence of each fracture type, relative incidences of each fracture type were calculated using the difference in the incidence in the RTA group compared to the non-RTA group. The association of mortality with the RTA group and RTA types was examined using a univariate binary logistic regression model. Multiple linear regression models were adjusted for significantly different cohort and hospital characteristics and used to compare differences between RTA and non-RTA patients in mean LOS and total charges. A subgroup analysis was conducted in the RTA group to assess differences in injury patterns, inpatient safety, and cost measures between different RTA types. The level of significance was set at two-sided P ≤ 0.05. All statistical analyses were performed on IBM Statistical Package for Social Sciences (Version 29.0.1.0).
Results
Cohort Characteristics
In total, there were 154,185 facial fracture patients identified during the 4-year study period, of which 16.9% (n = 26,115) were associated with RTAs and 83.1% (n = 128,070) were associated with other etiologies (Table 1). Patients in the RTA group commonly involved car occupants (40.7%), motorcycle riders (15.%), pedestrians (11.%), and bicycle riders (9.7%).
Table 1.
Cohort Characteristics for Overall Population in RTA vs Non-RTA Group.
| Cohort Characteristics | Overall | RTA | Non-RTA | P-Value |
|---|---|---|---|---|
| N | 154,185 (100) | 26,115 (16.9) | 128,070 (83.1) | N/A |
| Sex | ||||
| Female (%) | 45,490 (29.5) | 7700 (29.5) | 37,790 (29.5) | .968 |
| Male (%) | 108,650 (70.5) | 18,400 (70.5) | 90,250 (70.5) | – |
| Age (years) | ||||
| <18 (%) | 20,975 (13.6) | 3950 (15.1) | 17,025 (13.3) | <.001 |
| 18-24 (%) | 18,000 (11.7) | 4940 (18.9) | 13,060 (10.2) | <.001 |
| 25-44 (%) | 48,645 (31.4) | 9275 (35.5) | 39,190 (30.6) | <.001 |
| 45-54 (%) | 16,625 (10.8) | 3005 (11.5) | 13,620 (10.6) | <.001 |
| 55-64 (%) | 15,800 (10.2) | 2820 (10.8) | 12,980 (10.1) | .001 |
| ≥65 (%) | 34,305 (22.2) | 2115 (8.10) | 32,190 (25.1) | <.001 |
| Primary payer | ||||
| Medicare (%) | 35,985 (23.3) | 1805 (6.9) | 34,180 (26.8) | <.001 |
| Medicaid (%) | 44,950 (29.2) | 6925 (26.6) | 38,025 (29.8) | <.001 |
| Private (%) | 40,940 (26.6) | 11760 (45.2) | 29,180 (22.9) | <.001 |
| Self-pay (%) | 18,675 (12.1) | 3595 (13.8) | 15,080 (11.8) | <.001 |
| No charge (%) | 1800 (1.2) | 275 (1.1) | 1525 (1.2) | .060 |
| Other/unknown (%) | 11,365 (7.4) | 1660 (6.4) | 9705 (7.6) | <.001 |
| Discharge status | ||||
| Routine (%) | 114,200 (74.1) | 20,390 (78.1) | 93,810 (73.4) | <.001 |
| Transfer to short term hospital (%) | 2250 (1.5) | 470 (1.8) | 1780 (1.4) | <.001 |
| Transfer to other facility (%) | 19,735 (12.8) | 2650 (10.1) | 17,085 (13.4) | <.001 |
| Home health care (%) | 10,965 (7.1) | 1480 (5.7) | 9485 (7.4) | <.001 |
| Against medical advice (%) | 4450 (2.9) | 620 (2.4) | 3830 (3.0) | <.001 |
| Died in hospital (%) | 2340 (1.5) | 500 (1.9) | 1840 (1.4) | <.001 |
| Race | ||||
| White (%) | 83,930 (54.4) | 13,630 (54.4) | 70,300 (57.0) | <.001 |
| Black (%) | 29,980 (19.4) | 5030 (20.1) | 24,950 (22.0) | .556 |
| Hispanic (%) | 22,645 (14.7) | 4460 (17.8) | 18,185 (14.7) | <.001 |
| Asian or pacific islander (%) | 3520 (2.3) | 480 (1.9) | 3040 (2.5) | <.001 |
| Native american (%) | 1640 (1.1) | 215 (0.90) | 14,252 (1.2) | <.001 |
| Other (%) | 6730 (4.4) | 1260 (5.0) | 5470 (4.4) | <.001 |
| Median household income for patient zip code | ||||
| Q1: 0-25th percentile (%) | 52,540 (34.1) | 9000 (35.6) | 43,540 (35.3) | .363 |
| Q2: 26th-50th percentile (%) | 37,495 (24.3) | 6735 (26.60) | 30,760 (24.9) | <.001 |
| Q3: 51st-75th percentile (%) | 31,980 (20.7) | 5480 (21.6) | 26,500 (21.5) | .499 |
| Q4: 76th-100th percentile (%) | 26,810 (17.4) | 4100 (16.2) | 22,710 (18.4) | <.001 |
Compared to the non-RTA group, the RTA group was younger on average (35.5 vs 44.4 years, P < .001). There were more patients in the RTA group under the age of 44 (69.5% vs 54.1%, P < .001), with a third of all patients in the RTA cohort less than 24 years old (33.9% vs 23.5%, P < .001) (Figure 1). The RTA group had more patients who were covered by private insurance (45.2% vs 22.9%, P < .001), largely due to differences in Medicare coverage (6.9% vs 26.8%, P < 0.001). There were more Hispanic patients (17.8% vs 14.7%, P < .001) and less White (54.4% vs 57.0%, P < 0.001) and Asian patients (1.9% vs 2.5%, P < .001) in the RTA cohort. The RTA cohort had more patients in the bottom 50th percentile of median household income (58% vs 60.3%, P < .001).
Figure 1.
Age Distribution. The percentage of patients in the RTA vs non-RTA cohort distributed across different age groups.
While most facial fracture patients were residents in densely populated urban regions, the RTA group constituted a smaller proportion of patients compared to the non-RTA group (32.9% vs 36.4%, P < .001). The largest proportions of pedestrians and bicycle RTAs were found in urban areas, namely “central” (14.1% pedestrians, 12.0% bicycle) and “fringe” counties with over 1 million inhabitants (12.1% pedestrians, 10.6% bicycle). While the largest proportions of motorcycle RTAs were also found in central urban counties (14.5%), they were also evenly distributed across fringe areas (10.6%) and counties with populations ranging from 250,000-999,000 (9.1%). Car-related accidents were most prevalent in the latter (42.3%) and less populated micropolitan counties with less than 50,000 occupants (42.8%).
Patients in the RTA cohort were more frequently admitted to large (68.4% vs 65.3%, P < 0.001), public (20.4% vs 18.7%, P < 0.001), urban teaching hospitals (92.7% vs 90.6%, P < 0.001) compared to the those in the non-RTA group. The highest proportion of RTAs (24.1%) was localized to the South Atlantic region, including states like Maryland, Virginia, and Florida (Table 2). However, when examining the distribution of RTA types across regions (Figure 2), the Pacific (15.8%) and Mid-Atlantic regions (15.2%) had the largest proportions of pedestrians, while the Pacific (15.9%) and Mountain (12.7%)/Mid-Atlantic regions (12.1%) had the largest proportion of bicyclists. The largest proportion of motorcyclists was found in the Mid-Atlantic (19.8%) and Mountain (17.3%)/New England regions (17.2%) while the largest proportion of cars was found in the East South Central (46.6%) and East North Central (45.9%).
Table 2.
Hospital Characteristics for Overall Population in RTA vs Non-RTA Group. “Central” and “Fringe” Refers to Counties of Metro Areas of Greater Than or Equal to 1 Million Inhabitants.
| Hospital Characteristics | Overall | RTA | Non-RTA | P-Value |
|---|---|---|---|---|
| N | 154,185 (100) | 26,115 (16.9) | 128,070 (83.1) | N/A |
| Bed size | ||||
| Small (%) | 15,305 (9.9) | 2155 (8.3) | 13,150 (10.3) | <.001 |
| Medium (%) | 37,385 (24.2) | 6105 (23.4) | 31,280 (24.4) | <.001 |
| Large (%) | 101,495 (65.8) | 17,855 (68.4) | 83,640 (65.3) | <.001 |
| Location/Teaching status | ||||
| Rural (%) | 2570 (1.7) | 370 (1.4) | 2200 (1.7) | <.001 |
| Urban nonteaching (%) | 11,420 (7.4) | 1540 (6.0) | 9880 (7.7) | <.001 |
| Urban teaching (%) | 140,195 (90.9) | 24,205 (92.7) | 115,990 (90.6) | <.001 |
| Control | ||||
| Government, nonfederal (%) | 29,215 (18.9) | 5325 (20.4) | 23,890 (18.7) | <.001 |
| Private, not-for-profit (%) | 104,795 (68.0) | 17,305 (66.3) | 87,490 (68.3) | <.001 |
| Private, investor owned (%) | 20,175 (13.1) | 3485 (13.3) | 16,690 (13.0) | .172 |
| Hospital division | ||||
| New England (%) | 4990 (3.2) | 640 (2.5) | 4350 (3.4) | <.001 |
| Middle Atlantic (%) | 23,495 (15.2) | 3315 (12.7) | 20,180 (15.8) | <.001 |
| East North Central (%) | 19,835 (12.9) | 2930 (11.2) | 16,905 (13.2) | <.001 |
| West North Central (%) | 9900 (6.4) | 1760 (6.7) | 8140 (6.4) | .021 |
| South Atlantic (%) | 33,430 (21.7) | 6305 (24.1) | 27,125 (21.2) | <.001 |
| East South Central (%) | 9130 (5.9) | 2060 (7.9) | 7070 (5.5) | <.001 |
| West South Central (%) | 17,865 (11.6) | 3270 (12.5) | 14,595 (11.4) | <.001 |
| Mountain (%) | 11,450 (7.4) | 1935 (7.4) | 1935 (7.4) | .911 |
| Pacific (%) | 24,090 (15.6) | 20,190 (15.8) | 3900 (14.9) | <.001 |
| Population density | ||||
| Central | 54,045 (35.1) | 8450 (32.9) | 45,595 (36.4) | <.001 |
| Fringe | 31,865 (20.7) | 5400 (21.1) | 26,465 (21.1) | .825 |
| 250,000-999,000 | 31,360 (20.3) | 5530 (21.6) | 25,830 (20.6) | <.001 |
| 50-249,999 | 12,425 (8.1) | 1985 (7.7) | 10,440 (8.3) | .002 |
| Micropolitan | 12,300 (8.0) | 2385 (9.3) | 9915 (7.9) | <.001 |
| Non-metro/micropolitan | 8995 (5.8) | 1900 (7.4) | 7095 (5.7) | <.001 |
Figure 2.
Road Traffic Accident Type by Region. Pie graphs demonstrate the percentage of each RTA type across different census regions. Exploded pie slices represent the region where that RTA type is most prominently represented. West: Pacific (Washington, Oregon, California, Alaska, Hawaii) and Mountain (Montana, Idaho, Wyoming, Nevada, Utah, Colorado, Arizona, New Mexico). Northeast: Mid-Atlantic (New York, Pennsylvania, New Jersey) and New England (Maine, Vermont, New Hampshire, Massachusetts, Connecticut, Rhode Island). Midwest: West North Central (North Dakota, South Dakota, Nebraska, Kansas, Minnesota, Iowa, Missouri) and East North Central (Wisconsin, Illinois, Indiana, Michigan, Ohio). South: West South Central (Texas, Oklahoma, Arkansas, Louisiana), East South Central (Kentucky, Tennessee, Mississippi, Alabama), and South Atlantic (Maryland, Delaware, West Virginia, Virginia, North Carolina, South Carolina, Georgia, Florida).
Injury Patterns
Of the 26,115 patients in the RTA group, 34.8% (n = 9090) had isolated facial fractures while 65.2% (n = 17,025) combination fractures, or more than 1 facial fracture type concomitantly. The RTA group had a higher proportion of patients with combination fractures than the non-RTA group (65.2% vs 55.8%, P < .001). Within the RTA group, motorcyclists had the highest incidence of combination fractures (71%) (Table 3). Having an RTA-related facial fracture was associated with 50% increased odds of having a combination fracture compared to non-RTA etiologies (OR = 1.5, P < .001); the odds were highest for motorcyclists (OR = 1.8, P < .001). Over 1 third of the patients in the RTA group had 3 or more concomitant fractures compared to less than 1 fourth in the non-RTA group (36.1% vs 24.7%, P < .001) (Figure 3).
Table 3.
Injury Patterns, Inpatient Safety and Cost Measures for Facial Fractures by RTA Type.
| Pedestrian | Bicycle | Motorcycle | Car | P-Value | |
|---|---|---|---|---|---|
| N | 2880 | 2525 | 3910 | 10,640 | N/A |
| Number of facial fractures | |||||
| Isolated | 990 (34.4) | 760 (30.1) | 1135 (29.0) | 4015 (37.7) | <.001 |
| Combination | 1890 (65.6) | 1765 (69.9) | 2775 (71.0) | 6625 (62.3) | <.001 |
| Mean number of facial fractures | 2.3 | 2.4 | 2.5 | 2.2 | <.001 |
| Type of facial fracture | |||||
| Alveolus | 100 (3.5) | 80 (3.2) | 125 (3.2) | 345 (3.2) | .754 |
| Basilar | 90 (3.1) | 45 (1.8) | 75 (1.9) | 150 (1.4) | <.001 |
| Frontal | 560 (19.4) | 310 (12.3) | 515 (13.2) | 1455 (13.7) | <.001 |
| LeFort I | 60 (2.1) | 70 (2.8) | 170 (4.3) | 300 (2.8) | <.001 |
| LeFort II | 90 (3.1) | 60 (2.4) | 165 (4.2) | 185 (1.7) | <.001 |
| LeFort III | 45 (1.6) | 65 (2.6) | 110 (2.8) | 325 (3.1) | <.001 |
| Malar | 0 (0.0) | 0 (0.0) | 0 (0.0) | 10 (0.1) | <.001 |
| Mandibular | 570 (19.8) | 665 (26.3) | 915 (23.4) | 2415 (22.7) | <.001 |
| Maxillary | 225 (7.8) | 215 (8.5) | 270 (6.9) | 730 (6.9) | .008 |
| Nasal | 340 (11.8) | 250 (9.9) | 380 (9.7) | 1580 (14.8) | <.001 |
| Orbital | 290 (10.1) | 300 (11.9) | 460 (11.8) | 1730 (16.3) | <.001 |
| Zygomatic | 280 (9.7) | 370 (14.7) | 475 (12.1) | 790 (7.4) | <.001 |
| Unspecified | 230 (8.0) | 95 (3.8) | 250 (6.4) | 625 (5.9) | <.001 |
| Associated cervical spine injuries | |||||
| Overall | 320 (11.1) | 175 (6.9) | 445 (11.4) | 965 (9.1) | <.001 |
| Fracture | 300 (10.4) | 170 (6.7) | 400 (10.2) | 920 (8.6) | <.001 |
| Nerve | 40 (1.4) | 5 (0.2) | 75 (1.9) | 75 (0.7) | <.001 |
| Mean LOS, d | 7.8 | 4.0 | 5.5 | 5.1 | <.001 |
| Mean total charges, $ | $169,772 | $109,868 | $140,441 | 121,687 | <.001 |
| Mortality % | 100 (3.5) | 30 (1.2) | 105 (2.7) | 165 (1.6) | <.001 |
Figure 3.
Cumulative Fractures Distribution. The difference between the percentage of RTA and percentage of non-RTA patients that suffered 1 or multiple fractures is shown.
The incidence of each fracture type was compared between the RTA and non-RTA groups. Overall, the most common fracture types in the RTA group were mandibular (23.4%), orbital (14%), and frontal (13.7%) fractures (Table 4). However, when compared to the non-RTA group, the RTA group tended to involve more fractures in the midface rather than lower third fractures. The RTA group had higher rates of zygomatic (9.44% vs 6.82%, P < .001) and Lefort III (2.70% vs 0.83%, P < 0.001) fractures while the non-RTA group and markedly lower rates of mandibular fractures (23.4% vs 31.9%, P < .001) (Figure 4). The incidence of each fracture type for the different types of RTAs are shown (Table 3). In the RTA group, the 3 most common isolated fracture types were nasal (23.4%), frontal (22.3%), and mandibular (18.4%). The RTA group had more isolated nasal fractures (23.4% vs 20.4%, P < .001) and less frontal fractures (22.3% vs 27.0%, P < .001) than the non-RTA group. The 3 most common fracture types involved in combination fractures in the RTA group were mandibular (26.1%), orbital (14.3%), and zygomatic (12.0%). Compared to the non-RTA group, the combination fractures in the RTA group more frequently involved frontal fractures (9.16% vs 4.36%, P < .001) instead of mandibular fractures (26.1% vs 39.8%, P < .001).
Table 4.
Facial Fracture Type Incidence Rates Within RTA Group.
| Facial Fractures | Total | Isolated | Combination |
|---|---|---|---|
| N | 26,115 (100) | 9090 (34.8) | 17,025 (65.2) |
| Mandibular (%) | 6120 (23.4) | 1675 (18.4) | 4445 (26.1) |
| Orbital (%) | 3665 (14.0) | 1225 (13.5) | 2440 (14.3) |
| Frontal (%) | 3585 (13.7) | 2025 (22.3) | 1560 (9.2) |
| Nasal (%) | 3470 (13.3) | 2125 (23.4) | 1345 (7.9) |
| Zygomatic (%) | 2465 (9.44) | 430 (4.7) | 2035 (12.0) |
| Alveolus (%) | 870 (3.33) | 250 (2.8) | 620 (3.6) |
| LeFort I (%) | 800 (3.06) | 55 (0.6) | 745 (4.4) |
| LeFort III (%) | 705 (2.70) | 60 (0.7) | 645 (3.8) |
| LeFort II (%) | 630 (2.41) | 90 (1.0) | 540 (3.2) |
| Basilar (%) | 450 (1.7) | 205 (2.3) | 245 (1.4) |
| Maxillary (%) | 1835 (7.0) | 335 (3.7) | 1500 (8.8) |
| Malar (%) | 15 (0.06) | 5 (0.06) | 10 (0.06) |
| Unspecified (%) | 1505 (5.76) | 610 (6.7) | 895 (5.3) |
Figure 4.
Difference in fracture types in RTA vs non-RTA cohort for all, isolated, and combination facial fractures. The difference in incidence rate between the RTA and non-RTA group was calculated for each fracture type. Calculated differences are shown for each fracture type, sorted smallest to largest.
Associated cervical spinal injuries were identified in both the RTA and non-RTA cohort. The overall incidence of these injuries was 3.8% (n = 5865), including 5525 cervical spine fractures and 560 cervical nerve injuries. The RTA group exhibited a higher rate of associated cervical spinal injuries (9% vs 2.8%, P < .001), as well as elevated individual rates of fractures (8.5% vs 2.6%, P < .001) and nerve injuries (.9% vs .3%, P < .001). Having an RTA-related facial fracture was associated with 3.5 increased odds of having an associated cervical spinal injury (OR = 3.5, P < .001). Among RTA types, motorcyclists had the highest incidence of associated cervical spinal injuries (11.4%) and were most likely to have an associated cervical spinal injury with their facial fractures (OR = 3.4, P < .001).
Inpatient Safety and Cost Measures
While the overall inpatient mortality rate was low at 1.5% (n = 2340), the RTA group had a higher mortality rate than the non-RTA group (1.9% vs 1.4%, P < .001) (Table 5). Patients in the RTA group also had a longer mean length of stay (3.96 vs 5.27 days, P < .001) corresponding with higher mean total admission charges compared to the non-RTA group ($128,000 vs $79,400, P < .001). When stratified by RTA type, pedestrians had the greatest inpatient mortality (3.5%), mean length of stay (7.8 days), and mean admission charges ($170,000) (Table 3).
Table 5.
Inpatient Safety and Cost Measures for Facial Fractures by RTA (Univariate Analyses - T-Tests and Chi Square). LOS, Length of Stay. CI, Confidence Interval.
| Safety and Cost Measures | RTA | Non-RTA | P-Value |
| Mean LOS (days, 95% CI) | 5.27 | 3.96 | <.001 |
| Mean total charges ($, 95% CI) | 127,932 | 79,414 | <.001 |
| Mortality (%) | 1.9 | 1.4 | <.001 |
Univariate linear and logistic regression models were utilized to investigate the relationship of RTA-related facial fractures to inpatient safety and cost measures. The model revealed that facial fractures secondary to RTAs were associated with longer mean length of stay (B = 1.31, P < .001), admission charges (B = 48,518, P < .001), and mortality (OR = 1.34, P < .001) than those secondary to other causes. These associations remained significant for longer mean LOS (B = 1.13, P < .001) and admission charges (B = 38,881, P < .001) after adjusting for all variables that differed between the RTA and non-RTA groups in multivariate linear regression models (Table 6).
Table 6.
Inpatient Safety and Cost Measures for Facial Fractures by RTA (Multivariate Analyses - Multiple Linear Regression and Binary Logistic Regression). The Multivariate Logistic Regression Model was Unable to Handle the Number of Covariates. LOS, Length of Stay. CI, Confidence Interval.
| Safety and Cost Measures | OR/β | 95% CI | P-value |
|---|---|---|---|
| Mean LOS, d (β) | |||
| Unadjusted | 1.31 | 1.22-1.41 | <.001 |
| Adjusted | 1.13 | 1.03-1.22 | <.001 |
| Mean total charges, $ (β) | |||
| Unadjusted | 48,518 | 46,667-50,369 | <.001 |
| Adjusted | 38,881 | 37,035-40,726 | <.001 |
| Mortality % (OR) | |||
| Unadjusted | 1.34 | 1.21-1.48 | <.001 |
| Adjusted | - | - | - |
Discussion
The present study utilized a large representative sample of national inpatient data to characterize the burden of road traffic accidents on facial fractures in the United States. The present study demonstrates that, compared to other etiologies of facial fractures, RTA-related facial fractures disproportionately affect young adults, have a distinct injury pattern associated with multiple concomitant fractures and cervical spine injuries, and are associated with increased hospital costs and mortality. Notably, RTAs involving pedestrians had the greatest inpatient mortality, mean length of stay, and mean admission charges.
Our findings align with previous literature indicating that young adults are predominantly affected by RTA-related facial fractures,13,15,16 often attributed to non-compliance with seatbelts and intoxicated driving. 17 More recently, the increasing prevalence of smartphone usage has been linked to increasing RTAs for both drivers and pedestrians who become less aware of oncoming threats. 8 The pediatric population under 18 is another vulnerable group, 1 as an institutional study conducted in California found that pediatric facial trauma patients were caused by a higher percentage of motor vehicle accidents than adults (27% vs 13%). 18 The discrepancy in age and Medicare coverage between the RTA and non-RTA cohort is partially explained by the high incidence of falls leading to facial fractures in the elderly population.3,16
A notable aspect of the present analysis was the stratification of variables according to the 4 primary categories of road users: pedestrians, bicyclists, motorcyclists, and car occupants. The motivation behind this design was to characterize regional differences and injury patterns based on the RTA type. As observed in international cohort studies, geographical differences in traffic infrastructure strongly influences the incidence and type of RTA-related facial fractures. For instance, many Asian countries are known for widespread motorcycle use with inadequate sidewalk or roadway infrastructure. 19 Consequently, an international multicenter prospective study revealed that RTAs accounted for 68% of all facial fractures in the included Asian countries, largely involving motorcyclists. 12
The United States is unique in that there are significant regional differences in traffic infrastructure, resulting in varying traffic landscapes. The present study demonstrated that pedestrians and bicyclists were most injured in densely populated urban areas. Motorcyclists had the highest proportion in both densely populated areas and smaller counties whereas cars made up a higher proportion in rural counties. Given the prevalence of public transportation compared to car users in urban regions, 20 these findings align with the expected road user behaviors in these areas. For instance, many pedestrians and bicycle-related facial fractures occurred in the Mid-Atlantic region in the Northeast, which corresponds to the region with the highest reported commuters by transit (14.3%). 21
Patients in the RTA group had a greater risk of mortality, required longer hospital stays, and incurred greater hospital costs. Pedestrians were predominantly affected, likely due to their relative vulnerability on roads and lack of safety device protection. A potential explanation for these findings is due to the multiple concomitant facial fractures and cervical spine injuries, resulting in more complex and expensive management. A study utilizing a pediatric inpatient database demonstrated that fractures due to motor vehicle accidents and having 4 or more fracture concomitantly were both risk factors for longer hospital stay. 22 In addition, these patients may also suffer from other traumatic injuries not limited to the head and neck. While the present study only included patients with a primary facial fracture diagnosis, there were likely patients with other potentially serious traumatic injuries that contributed to these findings. Nonetheless, these findings indicate that craniofacial surgeons managing RTA patients must be prepared to handle multiple fractures and traumatic injuries simultaneously and coordinate within multidisciplinary teams to prioritize management.
This present study’s findings provide valuable insights for various stakeholders. For craniofacial surgeons managing patients with RTA-related fractures, the present study identifies distinct injury patterns associated with RTAs to help create tailored approaches for treatment. By highlighting trends in RTA types by geographical region and population density, providers and hospital administrators can have a more nuanced understanding of the traffic landscape surrounding their hospital, anticipate certain RTA types, and allocate resources accordingly. Finally, as RTAs are a preventable cause of injury and mortality, this study hopes to encourage craniofacial surgeons to advocate for better traffic infrastructure and policy to eliminate RTAs with the added benefit of decreasing hospital costs.
Conclusion
The present study demonstrates that, compared to other etiologies of facial fractures, RTA facial fractures disproportionately affect young adults, have a distinct injury pattern associated with multiple concomitant fractures and cervical spine injuries, and are associated with increased hospital costs and mortality. The present findings provide valuable insights, informing early involvement of craniofacial surgeons for the assessment of combination facial fractures and tailored treatment approaches for RTA patients.
Supplemental Material
Supplemental Material for The Burden of Road Traffic Accidents on Facial Fractures: National Trends, Injury Patterns, and Disparities in 154,185 Patients by Carol Wang, Daniel Y. Kwon, Olachi Oleru, Nargiz Seyidova, Peter E. Shamamian, Keisha Montalmant, Alex Sarosi, and Peter J. Taub in Craniomaxillofacial Trauma and Reconstruction
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding: The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the National Center for Advancing Translational Sciences (NCATS) TL1TR004420 NRSA TL1 Training Core in Transdisciplinary Clinical and Translational Science (CTSA).
Supplemental Material: Supplemental material for this article is available online.
Ethical Considerations
Ethical approval was not required for this study as it was not applicable.
ORCID iD
Carol Y. Wang https://orcid.org/0000-0002-3386-6260
References
- 1.Imahara SD, Hopper RA, Wang J, Rivara FP, Klein MB. Patterns and outcomes of pediatric facial fractures in the United States: A survey of the national trauma data bank. J Am Coll Surg. 2008;207(5):710-716. doi: 10.1016/j.jamcollsurg.2008.06.333 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Shayyab M, Alsoleihat F, Ryalat S, Khraisat A. Trends in the pattern of facial fractures in different countries of the world. Int J Morphol. 2012;30(2):745-756. doi: 10.4067/S0717-95022012000200065 [DOI] [Google Scholar]
- 3.Boffano P, Roccia F, Zavattero E, et al. European Maxillofacial Trauma (EURMAT) project: A multicentre and prospective study. J Cranio-Maxillo-Fac Surg. 2015;43(1):62-70. doi: 10.1016/j.jcms.2014.10.011 [DOI] [PubMed] [Google Scholar]
- 4.Allareddy V, Allareddy V, Nalliah RP. Epidemiology of facial fracture injuries. J Oral Maxillofac Surg. 2011;69(10):2613-2618. doi: 10.1016/j.joms.2011.02.057 [DOI] [PubMed] [Google Scholar]
- 5.VandeGriend ZP, Hashemi A, Shkoukani M. Changing trends in adult facial trauma epidemiology. J Craniofac Surg. 2015;26(1):108-112. doi: 10.1097/SCS.0000000000001299 [DOI] [PubMed] [Google Scholar]
- 6.Roccia F, Sotong J, Savoini M, Ramieri G, Zavattero E. Maxillofacial injuries due to traffic accidents. J Craniofac Surg. 2019;30(4):e288-e293. doi: 10.1097/SCS.0000000000005158. [DOI] [PubMed] [Google Scholar]
- 7.Choi SH, Gu JH, Kang DH. Analysis of traffic accident-related facial trauma. J Craniofac Surg. 2016;27(7):1682. doi: 10.1097/SCS.0000000000002916 [DOI] [PubMed] [Google Scholar]
- 8.Shaer M. Why are American drivers so deadly? The New York Times. https://www.nytimes.com/2024/01/10/magazine/dangerous-driving.html [Google Scholar]
- 9.Yellman MA. Motor vehicle crash deaths — United States and 28 other high-income countries, 2015 and 2019. MMWR Morb Mortal Wkly Rep. 2022;71. doi: 10.15585/mmwr.mm7126a1 [DOI] [PubMed] [Google Scholar]
- 10.Transportation USD of . Our Nation’s Roadway Safety Crisis. ArcGIS StoryMaps. 2023. https://storymaps.arcgis.com/stories/9e0e6b7397734c1387172bbc0001f29b [Google Scholar]
- 11.Injuries and Violence Are Leading Causes of Death | Injury Center. CDC. 2023. https://www.cdc.gov/injury/wisqars/animated-leading-causes.html [Google Scholar]
- 12.Romeo I, Roccia F, Aladelusi T, et al. A multicentric prospective study on maxillofacial trauma due to road traffic accidents: The world oral and maxillofacial trauma project. J Craniofac Surg. 2022;33(4):1057. doi: 10.1097/SCS.0000000000008440 [DOI] [PubMed] [Google Scholar]
- 13.Stacey DH, Doyle JF, Gutowski KA. Safety device use affects the incidence patterns of facial trauma in motor vehicle collisions: an analysis of the National Trauma Database from 2000 to 2004. Plast Reconstr Surg. 2008;121(6):2057-2064. doi: 10.1097/PRS.0b013e31817071fb [DOI] [PubMed] [Google Scholar]
- 14.Agency for Healthcare Research and Quality . HCUP-US NIS Overview. 2023. https://hcup-us.ahrq.gov/nisoverview.jsp [Google Scholar]
- 15.Chuang KT, Chu YY, Chen CT, Liao HT. Facial fracture patterns and severity of associated injuries after motorcycle accidents. Ann Plast Surg. 2023;90(1):S26. doi: 10.1097/SAP.0000000000003355 [DOI] [PubMed] [Google Scholar]
- 16.Irgebay Z, Kahan EH, Park KE, Choi J, Zellner EG. Characteristics and patterns of facial fractures in the elderly population in the United States based on trauma quality improvement project data. J Craniofac Surg. 2022;33(5):1294-1298. doi: 10.1097/SCS.0000000000008612. [DOI] [PubMed] [Google Scholar]
- 17.Hyman DA, Saha S, Nayar HS, Doyle JF, Agarwal SK, Chaiet SR. Patterns of facial fractures and protective device use in motor vehicle collisions from 2007 to 2012. JAMA Facial Plast Surg. 2016;18(6):455-461. doi: 10.1001/jamafacial.2016.0733 [DOI] [PubMed] [Google Scholar]
- 18.Su P, Paquet C, O'Dell K, et al. Trends in operative complex middle and upper maxillofacial trauma: A 17-year study. Laryngoscope. 2021;131(9):1985-1989. doi: 10.1002/lary.29442 [DOI] [PubMed] [Google Scholar]
- 19.World Health Organization . Global status report on road safety. 2018. https://www.who.int/publications-detail-redirect/9789241565684
- 20.McKenzie B. Who Drives to Work? Commuting by Automobile in the United States: 2013. Census.gov. 2023. https://www.census.gov/library/publications/2015/acs/acs-32.html [Google Scholar]
- 21.Burrows M, Burd C, McKenzie B. Commuting by Public Transportation in the United States; 2019. [Google Scholar]
- 22.Stanbouly D, Kocaelli H, Arce K, Selvi F. Risk factors for increased length of hospital stay among young pediatric patients who suffer facial fractures. Dent Traumatol. 2022;38(6):457-465. doi: 10.1111/edt.12776 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Supplemental Material for The Burden of Road Traffic Accidents on Facial Fractures: National Trends, Injury Patterns, and Disparities in 154,185 Patients by Carol Wang, Daniel Y. Kwon, Olachi Oleru, Nargiz Seyidova, Peter E. Shamamian, Keisha Montalmant, Alex Sarosi, and Peter J. Taub in Craniomaxillofacial Trauma and Reconstruction