Breed-specific Risk Factors and Clinical Outcomes in Pediatric Dog Bite Injuries: A 14-year Retrospective Analysis

Abstract

Background:

Dog bites are a major cause of pediatric injury, resulting in serious physical and psychological effects. This study examines breed-specific injury patterns, management, and outcomes.

Methods:

A retrospective review of pediatric dog bite cases from 2010 to 2024 was conducted. Variables included demographics, injury characteristics, dog breed, management, and outcomes (eg, infection, scarring, revisions). Descriptive and statistical analyses were performed.

Results:

Among 430 patients, the median age was 8.0 years (interquartile range 4.0–11.0). Bite frequency decreased with age (Spearman rho = −0.67; P = 0.001). Head and neck injuries were the most common (184 of 430, 42.8%). Specialist consultation occurred in 125 of 430 (29.1%), and 41 of 430 (9.5%) required operating room (OR) intervention. Head and neck injuries had higher rates of complex reconstruction (15 of 184, 8.2% versus 6 of 246, 2.4%; P = 0.007), soft-tissue loss (15.8% versus 7.7%; P = 0.009), and avulsion (19.6% versus 4.9%; P < 0.001) compared with injuries elsewhere. Among known breeds (n = 169), pit bulls were the most frequently reported (81 of 169, 47.9%). Management patterns and outcomes did not differ significantly by breed. In multivariable regression analysis (n = 169), dermal (OR 25.48, 95% confidence interval 2.97–679.02; P = 0.013) and subcutaneous involvement (OR 6.85, 95% confidence interval 1.33–54.54; P = 0.037) were associated with higher odds of abnormal scars, whereas breed was not. Patients requiring specialist consultation had higher return-to-OR rates than emergency department–managed patients (8.0% versus 0.0%; P < 0.001), with similar infection rates (6.4% versus 4.9%).

Conclusions:

Objective wound characteristics were more informative than breed classification for predicting management needs and scar outcomes. Prevention should prioritize safer child–dog interactions in familiar settings. Improved structured documentation may strengthen future risk stratification.

Takeaways

Question: Does reported dog breed correlate with injury severity and management in pediatric dog bites?

Findings: In this 14-year retrospective study of 430 pediatric dog bite cases, head and neck injuries were most common (42.8%). Breed was documented in 169 cases. Pit bulls were most often reported, but injury severity, management, and outcomes did not differ by breed. Dermal or subcutaneous involvement was associated with higher odds of abnormal scarring.

Meaning: Reported breed did not correlate with injury severity or management; objective wound characteristics should guide triage, counseling, and escalation of care.

INTRODUCTION

Dog bites represent a persistent public health concern and are among the leading causes of nonfatal injuries treated in emergency departments (EDs).^1,2 In the United States, approximately 4.5 million dog bites occur annually³; of these, 885,000 individuals seek medical care.^4–7 Pediatric patients comprise a substantial proportion of dog bite injuries. Children account for approximately 70% of all bite-related fatalities⁷ and are more likely than adults to require medical attention.⁴ Younger children are uniquely vulnerable due to their smaller stature, proximity to a dog’s face, and limited ability to recognize or avoid high-risk interactions.⁸ Such vulnerabilities can result in severe injuries, predominantly affecting the head and neck,^2,3,8 with significant physical disfigurement and psychosocial consequences.^9,10

Most dog bites occur in familiar environments, often involving known dogs in domestic settings.^2,8 In both children and adults, the presence of a dog in the household significantly increases the risk of bites, with the likelihood rising as the number of dogs in the home increases.⁴ Previous studies have investigated breed involvement and reported higher rates of severe injury in certain breeds,¹¹ raising important questions about breed-specific risk factors and the influence of environmental and situational variables.

This study expands upon existing literature by analyzing pediatric dog bite injuries at a single urban level I pediatric trauma center. Patterns in patient demographics, injury severity, and breed involvement are examined to identify factors influencing outcomes and to inform clinical management and prevention strategies. Additional outcomes are compared between cases involving specialist consultation and those managed exclusively by ED physicians.

METHODS

Study Design and Participants

This retrospective study reviewed electronic health records of pediatric patients (ages 0–21 y) who presented to our ED with dog bite injuries between January 2010 and July 2024. Patients were identified using International Classification of Diseases (ICD) codes (ICD-9 code E906.0 and ICD-10 code W54.0XXA). Injuries initially treated at outside hospitals were excluded. This study adhered to the Declaration of Helsinki and received institutional review board approval (STUDY660).

Variables and Data Collection

Data collection included demographics, injury characteristics, dog breed, patient–dog relationship, surgical interventions, and specialist involvement. Wound size was inconsistently documented in the ED and operative notes. When available, wound dimensions were defined as the largest documented linear measurement. When multiple injuries were present, the largest measurement was extracted. Tissue involvement was characterized by the deepest tissue layer exposed on examination. Complex reconstruction was defined as any repair requiring advanced techniques beyond simple primary closure (eg, local advancement flaps, skin grafts, composite grafts, cartilage grafts, or vascular repair such as arterial bypass with rotational muscle flap). Outcomes of interest included clinical assessment of infection, scarring at follow-up appointments, and need for scar revision procedures. Scar appearance was obtained from the latest documented follow-up.

Statistical Analysis

Patient demographics and dog bite characteristics were summarized using descriptive statistics. The trend in dog bite case counts by age was assessed using Spearman rank correlation. Mann-Whitney U or Kruskal–Wallis tests were used for continuous variables, and χ² or Fisher exact tests were used for categorical variables. Breed-stratified univariate comparisons and multivariable analysis, including dog breed, were limited to cases with documented breed information. Multivariable logistic regression was used to evaluate abnormal scar outcomes (hypopigmented, hypertrophic, and keloid), while adjusting for patient age, dog breed, anatomical location, wound dimension, and tissue involvement. For multivariable analysis, patients reporting multiple races were included in the “other” category, and Rottweilers and huskies were combined with the “other” breed category. Analyses were conducted using R software (version 4.5.1).

RESULTS

A total of 430 pediatric patients with dog bite injuries were included. The median age was 8.0 years (interquartile range [IQR] 4.0–11.0 y). Dog bite case frequency decreased with age (Spearman rho = −0.67; P = 0.001), with younger age groups accounting for a higher proportion of cases (Fig. 1). Dog bite injuries were more common among male patients (55.3%). Most patients identified as Black or African American (62.3%), followed by White (17.2%) and other races (16.0%). Overall, 35.8% of patients identified as not Hispanic, Latino, or of Spanish origin; identified as 16.0% as Hispanic or Latino; and 48.1% did not report their ethnicity (Table 1).

Fig. 1.

Dog bite frequency by patient age and sex (n = 430).

Table 1.

Demographics and Characteristics of Pediatric Patients Who Presented to the ED With a Dog Bite Injury Between 2010 and 2024

Characteristics	Frequency, n (%)
No. patients	430
Age, y
Median (IQR)	8.0 (4.0–11.0)
Sex
Male	238 (55.3)
Female	192 (44.7)
Race
Black/African American	268 (62.3)
White	74 (17.2)
Other	69 (16.0)
Multiple	3 (0.7)
Not provided	16 (3.7)
Ethnicity
Hispanic/Latino	69 (16.0)
Not Hispanic/Latino	154 (35.8)
Not provided	207 (48.1)
Insurance
Public	302 (70.2)
Private	111 (25.8)
Self-pay	17 (4.0)

Most dog bite incidents involved family-owned pets (50.8%) or domesticated dogs owned by an individual unrelated to the patient (44.3%). A minority (4.9%) was caused by stray dogs or dogs without a known handler. Most patients (62.5%) were transported to the ED by a family member, whereas 37.4% arrived by ambulance.

Injury Characteristics and Intervention

Dog bite injuries were stratified by bite location and injury severity. The most common injury site was the head and neck (42.8%), followed by the hand (17.0%), lower extremity (13.5%), arm (12.3%), multiple sites (8.8%), and other locations (5.6%). Among the 253 of 430 patients with documented wound dimensions, the median wound dimension was 2.0 cm (IQR 1.0–3.0 cm). Injury was limited to the epidermis in 60.2% of cases, involved the dermis in 12.8%, and extended into the subcutaneous layer in 27.0% (Table 2).

Table 2.

Characteristics of Dog Bite Injuries

Characteristics	Frequency, n (%)
N	430
Anatomical location
Head/neck	184 (42.8)
Upper extremity—arm	53 (12.3)
Upper extremity—hand	73 (17.0)
Lower extremity	58 (13.5)
Multiple	38 (8.8)
Other	24 (5.6)
Wound dimension,* cm
Median (IQR)	2.0 (1.0–3.0)
Tissue involvement
Epidermis	259 (60.2)
Dermis	55 (12.8)
Subcutaneous	116 (27.0)
Soft-tissue loss	48 (11.2)
Avulsion	48 (11.2)
Fracture	17 (4.0)
Specialist consult	125 (29.1)
OR visit	41 (9.5)
Intervention
Irrigation only	251 (58.4)
Irrigation and primary closure	140 (32.5)
Debridement and primary closure	18 (4.2)
Complex reconstruction	21 (4.9)

^*Excludes n = 177 missing wound measurements.

Among head and neck injuries (n = 184), injuries involving multiple head and neck subsites were the most frequent (50 of 184, 27.2%), followed by isolated injuries to the lips (41 of 184, 22.3%) and cheek (31 of 184, 16.8%). Injuries to the head and neck were significantly more likely to require complex reconstruction than injuries elsewhere (15 of 184, 8.2% versus 6 of 246, 2.4%; P = 0.007). Head and neck injuries also had significantly higher rates of soft-tissue loss (29 of 184, 15.8% versus 19 of 246, 7.7%; P = 0.009) and avulsion (36 of 184, 19.6% versus 12 of 246, 4.9%; P < 0.001) compared with injuries to other regions. Similarly, injuries spanning multiple anatomical locations had higher rates of soft-tissue loss (10 of 38, 26.3% versus 38 of 392, 9.7%; P = 0.002) and avulsion (9 of 38, 23.7% versus 39 of 392, 9.9%; P = 0.010) than single-region injuries.

Most injuries were managed with irrigation alone (58.4%) or irrigation with primary closure (32.5%). Debridement with primary closure (4.2%) and complex reconstruction (4.9%) accounted for the remainder (Fig. 2). Specialists, mainly plastic and reconstructive surgeons, were consulted in 29.1% of cases (Fig. 3), and 9.5% required operating room (OR) intervention (Table 2).

Fig. 2.

Distribution of intervention types across different anatomical locations of dog bite injuries.

Fig. 3.

Specialties consulted for dog bite injuries (N = 125).

Dog Breed and Breed-stratified Outcomes

Breed was unknown in 60.7% (261 of 430) of cases. For completeness, Supplemental Digital Content 1 shows the full cohort distribution, including the unknown breed category. (See table, Supplemental Digital Content 1, which displays the injury characteristics, management, and outcomes by dog breed, including the unknown breed category, https://links.lww.com/PRSGO/E985.) Among cases with documented breed (169 of 430), pit bulls were most common (81 of 169, 47.9%), followed by German shepherds (17 of 169, 10.1%), Rottweilers (8 of 169, 4.7%), and huskies (8 of 169, 4.7%). In the entire cohort (N = 430), pit bulls accounted for 18.8% of cases; German shepherds, 4.0%, and Rottweilers and huskies, 1.9% each (Fig. 4). Less common breeds were grouped as “other” (55 of 430, 12.8%), including mixed breeds, Chihuahuas, bulldogs, Labrador retrievers, Yorkshire terriers, golden retrievers, cane corsos, Doberman pinschers, Great Danes, poodles, Akitas, Belgian Malinois, chow chows, dachshunds, Pomeranians, pugs, and shih tzus.

Fig. 4.

Distribution of dog bite injuries across breeds. *Other breeds category includes a range of less commonly involved breeds, as defined in the Results section.

In univariate analysis among cases with known dog breed, head and neck injuries were more common in cases involving pit bulls (n = 28, 34.6%) and Rottweilers (n = 4, 50.0%). In contrast, German shepherd and husky bites most often involved the hand (n = 5, 29.4%) and lower extremity (n = 4, 50.0%), respectively (Fig. 5). Median wound dimension did not differ significantly by breed (P = 0.150), although husky bites had the largest median wound dimension (3.5 cm, IQR 2.8–4.3 cm). Depth of tissue involvement and rates of soft-tissue loss, avulsion, and fracture did not differ significantly across breeds (Table 3). Management patterns, including specialist consultation, OR intervention, and type of intervention, did not differ significantly across breeds. Irrigation alone was the most common intervention for all breeds, ranging from 54.5% to 75.0%. Univariate analysis showed no significant differences in return to the OR, infection, length of stay (LOS), and scar type distribution by breed. Hypopigmented scars were the most common scar type across all breeds. Pit bull injuries had the most diverse scar presentation, with 25.9% (21 of 81) normal-appearing scars, 70.4% (57 of 81) hypopigmented scars, 1.2% (1 of 81) hypertrophic scars, and 2.5% (2 of 81) keloids.

Fig. 5.

Distribution of dog bite injuries by anatomical location and breed. *Other breeds category includes a range of less commonly involved breeds, as defined in the Results section.

Table 3.

Injury Characteristics, Management, and Outcomes by Known Dog Breed

Characteristics	Level	Pit Bull	German Shepherd	Rottweiler	Husky	Other*	P
N		81	17	8	8	55
Anatomical location, n (%)	Head/neck	28 (34.6)	4 (23.5)	4 (50.0)	1 (12.5)	21 (38.2)	0.819
	Upper extremity—arm	8 (9.9)	2 (11.8)	0 (0.0)	0 (0.0)	6 (10.9)
	Upper extremity—hand	13 (16.0)	5 (29.4)	1 (12.5)	2 (25.0)	6 (10.9)
	Lower extremity—leg/foot	19 (23.5)	3 (17.6)	2 (25.0)	4 (50.0)	10 (18.2)
	Other	6 (7.4)	1 (5.9)	0 (0.0)	0 (0.0)	2 (3.6)
	Multiple	7 (8.6)	2 (11.8)	1 (12.5)	1 (12.5)	10 (18.2)
Wound dimension, cm	Median (IQR)	2.0 (1.0–3.0)	1.4 (1.0–3.4)	0.8 (0.4–1.3)	3.5 (2.8–4.3)	2.0 (1.0–2.8)	0.150
Tissue involvement, n (%)	Epidermis	50 (61.7)	6 (35.3)	5 (62.5)	5 (62.5)	31 (56.4)	0.732
	Dermis	8 (9.9)	3 (17.6)	1 (12.5)	1 (12.5)	6 (10.9)
	Subcutaneous	23 (28.4)	8 (47.1)	2 (25.0)	2 (25.0)	18 (32.7)
Soft-tissue loss, n (%)		13 (16.0)	3 (17.6)	1 (12.5)	1 (12.5)	9 (16.4)	>0.999
Avulsion, n (%)		9 (11.1)	4 (23.5)	2 (25.0)	1 (12.5)	5 (9.1)	0.324
Fracture, n (%)		2 (2.5)	1 (5.9)	0 (0.0)	1 (12.5)	3 (5.5)	0.395
Specialist consult, n (%)		22 (27.2)	5 (29.4)	2 (25.0)	2 (25.0)	16 (29.1)	0.997
OR visit, n (%)		9 (11.1)	3 (17.6)	1 (12.5)	1 (12.5)	8 (14.5)	0.909
Intervention, n (%)	Irrigation only	50 (61.7)	11 (64.7)	6 (75.0)	5 (62.5)	30 (54.5)	0.847
	Irrigation and primary closure	23 (28.4)	3 (17.6)	1 (12.5)	2 (25.0)	17 (30.9)
	Debridement and primary closure	2 (2.5)	2 (11.8)	0 (0.0)	0 (0.0)	3 (5.5)
	Complex reconstruction	6 (7.4)	1 (5.9)	1 (12.5)	1 (12.5)	5 (9.1)
Return to OR, n (%)		4 (4.9)	1 (5.9)	0 (0.0)	1 (12.5)	1 (1.8)	0.450
Infection, n (%)		5 (6.2)	0 (0.0)	0 (0.0)	1 (12.5)	1 (1.8)	0.347
LOS, d	Median (IQR)	0.0 (0.0–0.0)	0.0 (0.0–1.0)	0.0 (0.0–1.0)	0.0 (0.0–0.0)	0.0 (0.0–0.0)	0.381
Scar type, n (%)	Normal	21 (25.9)	3 (17.6)	3 (37.5)	0 (0.0)	20 (36.4)	0.384
	Hypopigmented	57 (70.4)	13 (76.5)	5 (62.5)	8 (100.0)	34 (61.8)
	Hypertrophic	1 (1.2)	1 (5.9)	0 (0.0)	0 (0.0)	1 (1.8)
	Keloid	2 (2.5)	0 (0.0)	0 (0.0)	0 (0.0)	0 (0.0)

^*Other documented breeds include mixed breeds (16), Chihuahuas (6), bulldogs (6), Labrador retrievers (4), Yorkshire terriers (4), golden retrievers (3), cane corsos (3), Doberman pinschers (2), Great Danes (2), poodles (2), Akitas (1), Belgian Malinois (1), chow chows (1), dachshunds (1), Pomeranians (1), pugs (1), and shih tzus (1).

Multivariable logistic regression evaluating abnormal scar outcomes (hypopigmented, hypertrophic, or keloid versus normal) was performed among cases with documented breed (n = 169). Injuries involving the dermis (OR 25.48, 95% confidence interval 2.97–679.02; P = 0.013) and subcutaneous tissue (OR 6.85, 95% confidence interval 1.33–54.54; P = 0.037) were associated with higher odds of abnormal scars compared with injuries confined to the epidermis. Dog breed, anatomical location, age, race, and wound dimension were not significantly associated with abnormal scar outcomes (Table 4).

Table 4.

Multivariable Logistic Regression for Abnormal Scar Outcomes (N = 169)

Characteristics	OR (95% CI)	P
Dog breed
Pit bull	Ref.
German shepherd	0.96 (0.13–9.81)	0.969
Other	0.70 (0.20–2.45)	0.581
Anatomical location
Head/neck	Ref.
UE—arm	2.86 (0.42–26.86)	0.305
UE—hand	0.55 (0.08–3.52)	0.517
Lower extremity	1.75 (0.35–9.43)	0.500
Multiple	1.22 (0.13–14.32)	0.861
Other	0.40 (0.02–5.30)	0.503
Age, y	0.81 (0.67–1.95)	0.135
Wound dimension, cm	1.23 (0.85–2.11)	0.402
Tissue involvement
Epidermis	Ref.
Dermis	25.48 (2.97–679.02)	0.013
Subcutaneous	6.85 (1.33–54.54)	0.037

Rottweilers and huskies were combined with the other breed category. Bold values indicate statistical significance (P < 0.05).

CI, confidence interval; OR, odds ratio; UE, upper extremity.

Outcomes by Specialist Versus ED Management

Outcomes were compared between cases with specialist consultation (n = 125) and those managed exclusively by ED physicians (n = 305). No patients managed solely by ED physicians required return to the OR after initial treatment, whereas 8% (10 of 125) of patients with specialist consultation required a subsequent return to the OR (P < 0.001). Infection rates were comparable between groups (4.9% in ED-managed versus 6.4% in specialist-managed, P = 0.535). Although median LOS was 0.0 days in both groups, patients treated by a specialist had a slightly longer LOS (IQR 0.0–1.0 d) compared with the ED-managed group (IQR 0.0–0.0 d; P < 0.001), suggesting that a greater proportion of specialist consult patients required overnight observation or admission. Because specialist consultation is typically obtained for more complex injuries, the unadjusted LOS difference likely reflects differences in injury complexity rather than physician specialty. Univariate analysis showed that scar type distribution also differed significantly between specialist and ED management (P < 0.001) (Table 5). Of note, 1 patient not seen by a specialist developed a hematoma, and another seen by a plastic surgeon experienced wound dehiscence.

Table 5.

Outcomes by Specialist Consult Versus ED Management

Characteristics	Specialist	ED	P
N	125	305
Return to OR, n (%)	10 (8.0)	0 (0.0)	<0.001
Infection, n (%)	8 (6.4)	15 (4.9)	0.535
LOS
Median (IQR)	0.0 (0.0–1.0)	0.0 (0.0–0.0)	<0.001
Scar type, n (%)
Normal  Hypopigmented  Hypertrophic  Keloid	16 (12.8) 98 (78.4) 9 (7.2) 2 (1.6)	107 (35.1) 198 (64.9) 0 (0.0) 0 (0.0)	<0.001

Bold values indicate statistical significance (P < 0.05).

Antibiotic Use

Among the 430 patients, 359 (83.5%) received antibiotics in the ED. Amoxicillin–clavulanate was most frequently prescribed (n = 307), followed by ampicillin–sulbactam (n = 28), clindamycin (n = 13), amoxicillin (n = 6), ceftriaxone (n = 6), metronidazole (n = 1), cefazolin (n = 1), and doxycycline (n = 1). Antibiotics were more often used for bites with higher risk features, such as head and neck or hand/near-joint involvement and deeper or complex wounds. Some patients received more than 1 type of antibiotic, so the total count exceeds the number of individual cases.

DISCUSSION

Dog bite injuries are a common cause of pediatric ED visits and can impose significant physical and psychological consequences.^3,10,12 Consistent with prior pediatric series,^13–16 we found in our 14-year analysis that dog bite injuries were concentrated in younger children and decreased in frequency with age. Prior studies suggest that shorter stature, limited situational awareness, and reduced motor control or strength increase younger children’s risk of bite injuries during close interactions with dogs.^8,13,16 Most of our cases occurred in familiar settings with a known dog handler. Most wounds were managed in the ED; however, nearly one-third involved specialist consultation, and roughly 1 in 10 required OR intervention. Compared with other locations, dog bites to the head and neck were associated with higher rates of soft-tissue loss and avulsion and were more likely to require complex reconstruction. The proximity of the head and neck to the dog’s mouth in younger children can also explain how bites more commonly affect the head and neck in this age group.^2,12,13 Bites to multiple anatomical locations had higher rates of soft-tissue loss and avulsion than bites to a single location. Deeper tissue involvement was independently associated with abnormal scar outcomes compared with injuries confined to the epidermis after adjustment.

It is well documented in the literature that the distribution of reported breeds among dog bite injuries varies by geographic region, with pit bulls or pit bull mixes often identified as the most common breed involved in pediatric injuries.^{11,15,17–20} Our data further support this observation. Some studies have also reported an association between pit bulls and more severe injuries on average.^2,13,21 However, breed is often not documented, and when known, it is based on victim or witness report,¹¹ raising concern about bias toward the subset for which breed is documented. Further, the “pit bull” breed label is variably defined across studies and may reflect a common phenotype rather than objective verification.^21,22 Consequently, pit bull–type dogs are often profiled as dangerous and stigmatized, and breed-specific legislation has been enacted to ban ownership in certain places.²³ Given geographic variability in breed prevalence and dog bite referral patterns, these reported trends should not be interpreted as true population-based risk.

In our cohort, head and neck injuries were more common in cases attributed to pit bulls, and hand and lower extremity injuries were more common among German shepherds and huskies; yet, we did not identify breed-based differences in injury severity, management patterns, or outcomes. In adjusted analysis, breed was not independently associated with abnormal scar outcomes either. Our findings therefore caution against using breed to influence resource use and reinforce that wound characteristics and patient factors remain primary drivers for triage, counseling, and management decisions.

Specialists were consulted in nearly a third of cases in our hospital, but OR intervention was required in a smaller subset, indicating that many consultations resulted in bedside evaluation and management without proceeding to the OR. Patients managed with specialist consultation had a slightly longer LOS and were the only group to require subsequent returns to the OR. The distribution of scar outcomes also differed between specialist-consulted and ED-managed patients. However, because comparisons were unadjusted and specialist consultation for dog bite injuries is typically sought for more complex wounds, the differences in LOS, scar outcomes, and return to the OR likely reflect baseline injury severity rather than the effect of the specialty involved. Clinically, these findings support escalation of care based on wound characteristics―including injuries involving critical structures, suspected fracture or neurovascular compromise, deep tissue involvement, head and neck location, or multisite injury―whereas superficial, single-region injuries without these features can often be managed in the ED with irrigation with or without primary closure and follow-up.¹⁶

Current protocol for dog bites at our facility includes notification of animal control or public health authorities in the jurisdiction where the bite occurred to facilitate locating and monitoring the dog and rabies postexposure prophylaxis (PEP) if the animal is unknown or cannot be located. We follow Centers for Disease Control and Prevention–based guidelines for rabies PEP.²⁴ PEP is typically deferred when the dog is known and available for observation or is being actively located by animal control. When the dog cannot be located or observed, the rabies vaccine is administered with rabies immunoglobulin when indicated per protocol. This regimen is adjusted for previously vaccinated and immunocompromised patients.

Per our institutional protocol, antibiotic prophylaxis is generally reserved when infection risk is higher (eg, hand/near-bone or joint involvement, face injuries, deep puncture or crush injury, immunocompromise, delayed presentation [>6–12 h for extremity wounds], and wounds requiring primary closure or OR repair). Antibiotics were frequently prescribed in our cohort (n = 359, 83.5%), and overall infection rates were low (n = 23, 5.3%). Admission is considered for delayed presentation with cellulitis, bone or joint involvement with concern for compartment syndrome, inability to tolerate PO medication, or wounds requiring OR repair. Although amoxicillin–clavulanate accounted for most antibiotic use, other antibiotics likely reflect variation in allergy history, parenteral or oral therapy needs, and practice-pattern variation across the extended study period. In our current institutional order set, amoxicillin–clavulanate is first-line oral therapy. For confirmed penicillin allergy, trimethoprim–sulfamethoxazole and clindamycin are recommended. When intravenous therapy is warranted, ampicillin–sulbactam is recommended with ceftriaxone and clindamycin or trimethoprim–sulfamethoxazole and clindamycin as allergy-based alternatives.

Given that the vast majority of incidents in our cohort involved a dog with a known handler (50.8% family-owned pets and 44.3% owned by an unrelated individual), prevention efforts must emphasize adult responsibility, as young children are neither able to influence decisions on dog ownership or selection nor developmentally equipped to recognize or avoid high-risk interactions. Similar to the findings of this study, prior studies have found that pediatric dog bites commonly occur at home and often involve familiar dogs.^{2,11,14,18,25} Initiatives to educate parents, caregivers, and children (in an age-specific manner) can improve knowledge of safer interactions and understanding of dog behavior. However, education does not necessarily translate into consistent behavior change, and very young children may not be able to apply safety rules reliably.^25–27 Therefore, multiple approaches, including (1) parent and caregiver education on canine stress and bite-risk warning signals, (2) proactive supervision of young children even around familiar dogs, (3) responsible ownership practices such as early dog socialization and consistent dog training practices, and (4) community outreach through social media platforms to increase awareness about dog bite prevention, may be more effective than any single intervention alone.^2,26 Anticipatory guidance should emphasize active supervision of young children around any dog and avoidance of face-to-face contact, given the predominance of head and neck injuries in younger children.¹⁸ Families should also limit interactions during predictable high-risk situations (eg, when a dog is eating, resource guarding, sleeping, or showing signs of fear/stress).^2,18,26 When direct supervision is not possible, controlling the environment, such as separating dogs and children using physical barriers, may reduce risks. If consistent supervision and separation are not realistically feasible, families should consider delaying the introduction of a dog until children are older.

This study has limitations. Its retrospective design relied on clinical documentation, and key variables, such as dog breed and wound measurements, were inconsistently recorded. When documented, wound dimension was extracted as the largest linear measurement, which may underestimate total wound burden in patients with multiple wounds. Similar to many other studies on pediatric dog bite injuries,^2,18,19,21 breed was unavailable for many encounters in our cohort, and known breeds were based on report rather than objective verification. This introduces bias and may weaken breed-stratified comparisons. Several breed categories also included a few cases, which limited statistical power to detect differences and precluded stable adjusted estimates for some breeds. Because ethnicity data were missing in nearly half of patients, we excluded race/ethnicity as an independent predictor in multivariable scar analysis despite studies showing differences in scar outcomes across race/ethnicity. Further, scar appearance documented during follow-up visits is subject to bias based on the individual physician’s assessment. Many patients lacked uniform long-term follow-up, which may have led to undercapture of scar maturation and delayed complications. This was a single-center study at an urban level I pediatric trauma center, and we excluded patients initially treated at outside hospitals due to incomplete data. As a result, findings may not generalize to other settings. Future prospective studies using structured ED documentation templates may improve data quality and capture injury circumstances and standardized wound features, strengthening comparisons between breed and outcomes. Longitudinal studies assessing functional, aesthetic, and psychosocial outcomes after intervention may provide further insight into long-term care needs.

CONCLUSIONS

Pediatric dog bite injuries remain a significant public health concern and can impose substantial physical, emotional, and financial consequences. In this 14-year retrospective review, injuries were more common in younger children and most frequently involved the head and neck. Head and neck and multisite injuries were associated with higher rates of soft-tissue loss and avulsion. Head and neck injuries were more likely to require complex reconstruction. Deeper tissue involvement was the strongest predictor of abnormal scar appearance. Overall, these findings support prioritizing wound characteristics and patient factors over breed-specific risk in triage, counseling, and management decisions.

DISCLOSURE

The authors have no financial interest to declare in relation to the content of this article.

ACKNOWLEDGMENT

The authors thank Md Sohel Rana for his guidance with statistical analysis.

ETHICAL APPROVAL

This study was approved by the institutional review board for human research at the Children’s National Hospital, Washington, DC.

Supplementary Material

gox-14-e7856-s001.pdf ^{(188KB, pdf)}