This study investigated characteristics of 809 children presenting to the emergency department for epistaxis (nosebleed) between 2013-2022. The majority (92.7%) were treated medically with nasal compression or intranasal medications. Older age, bleeding duration over 30 minutes, and antiplatelet medication use were associated with the 6.3% who received procedural intervention like cauterization. Procedural control led to higher rates of transfusion and admission. The results provide guidance on pediatric epistaxis patients needing emergency referral.
Acute Management of Pediatric Epistaxis in the Emergency Department
1. Acute Management of Pediatric
Epistaxis in the Emergency
Department
2023 Barsan Emergency Medicine Research Forum
Andrew Shieh, MD
Pediatric Emergency Medicine Fellow
Sarah Tomlinson, MD
Clinical Assistant Professor, Departments of Emergency Medicine and Pediatrics
3. Introduction
• Epistaxis is a common cause of PED visits.
• Pediatric nosebleeds are seldomly serious, regardless of risk factors,
and are usually treated conservatively.
• The estimated proportion of pediatric epistaxis requiring procedures
to control bleeding is 7%.
4. Aim
• To investigate the demographics of children presenting to a PED for
epistaxis and identify patient characteristics associated with
procedural intervention
5. Hypothesis
• In patients who present to a PED for epistaxis, there is no difference
in characteristics between those who are treated medically and those
requiring procedures
6. Methods
• We performed a multivariable logistic regression analysis of different
characteristics to determine which patients require procedures
Medical Management Procedural Control
8. Inclusion/
Exclusion
297,103 ED Visits
Between 2013-2022
858 ED Visits With Confirmed Epistaxis
49 Visits excluded:
33 Post-op otolaryngology procedure within 30 days
2 Nasopharyngeal mass/Angiofibroma
3 Nasal arteriovenous malformation
11 Visits from same patient
809 Patients with ED Visit for
Epistaxis
296,245 ED Visits Without
Epistaxis
9. Total Cohort N=809 Medical Management n=758 Procedural Control n=51 P value
Age (Years)
Mean 9.30 9.08 12.61 <0.001
Median 8.66 8.21 13.89 <0.001
Sex 0.57
Male 443 (55%) 417 (55%) 26 (51%)
Female 366 (45%) 341 (45%) 25 (49%)
Race 0.21
White/Caucasian 481 (59%) 449 (59%) 32 (63%)
Black/African
American
168 (21%) 159 (21%) 9 (17.5%)
Asian 65 (8%) 64 (8.5%) 1 (2%)
Other 95 (12%) 86 (11.5%) 9 (17.5%)
Patient Characteristics
10. Total Cohort N=809 Medical Management n=758 Procedural Control n=51 P value
Duration of Nosebleed <0.001
Under 30
Minutes
507 (63%) 500 (66%) 7 (14%)
30 Minutes or
Longer
275 (34%) 234 (31%) 41 (80%)
No Data 27 (3%) 24 (3%) 3 (6%)
Laterality 0.95
Unilateral 480 (59%) 445 (59%) 35 (69%)
Bilateral 223 (28%) 207 (27%) 16 (31%)
No Data 106 (13%) 106 (14%) 0 (0%)
Epistaxis Characteristics
13. Association Model of Epistaxis Risk
Factors for Procedural Control
Adjusted OR (95% CI) Adjusted P value
Older Age 1.09 (1.02-1.16) 0.01
Duration ≥ 30 Minutes 7.19 (3.05-16.93) <0.001
14. Association Model of Epistaxis Risk
Factors for Procedural Control
Adjusted OR (95% CI) Adjusted P value
Older Age 1.09 (1.02-1.16) 0.01
Duration ≥ 30 Minutes 7.19 (3.05-16.93) <0.001
Risk Factors
Respiratory Viral Illness 0.60 (0.27-1.34) 0.21
Direct Trauma 0.45 (0.14-1.45) 0.18
Allergic Rhinitis 0.68 (0.26-1.78) 0.43
Digital Trauma 1.29 (0.40-4.13) 0.66
Known Bleeding Disorder 2.10 (0.84-5.24) 0.11
Intranasal Corticosteroids 1.79 (0.43-7.50) 0.42
Antiplatelet Medication 7.70 (2.59-22.88) <0.001
Foreign Body 1.40 (0.15-12.68) 0.76
Known Oncological Condition 3.05 (0.86-10.76) 0.08
Anticoagulation Medication 0.80 (0.12-5.30) 0.81
18. Conclusion
6.3% of children with epistaxis receive procedures in the ED.
Children with older age, prolonged duration of bleeding, and antiplatelet
medication use are at risk for procedural control.
Procedural control of pediatric epistaxis is associated with transfusions
and hospital admission.
This project provides primary care physicians additional knowledge and
data regarding which children require referral to the ED for epistaxis.
19. Conclusion
6.3% of children with epistaxis receive procedures in the ED.
Children with older age, prolonged duration of bleeding, and antiplatelet
medication use are at risk for procedural control.
Procedural control of pediatric epistaxis is associated with transfusions
and hospital admission.
This project provides primary care physicians additional knowledge
regarding which children require referral to the ED for epistaxis.
21. References
• Patel N, Maddalozzo J, Billings KR. An update on management of
pediatric epistaxis. Int J Pediatr Otorhinolaryngol. 2014 Aug;78(8):1400-4.
doi: 10.1016/j.ijporl.2014.06.009. Epub 2014 Jun 16. PMID: 24972938.
• Shay S, Shapiro NL, Bhattacharyya N. Epidemiological characteristics of
pediatric epistaxis presenting to the emergency department. Int J Pediatr
Otorhinolaryngol. 2017 Dec;103:121-124. doi:
10.1016/j.ijporl.2017.10.026. Epub 2017 Oct 17. Erratum in: Int J Pediatr
Otorhinolaryngol. 2018 Dec;115:193. PMID: 29224751.
• Baugh TP, Chang CWD. Epidemiology and Management of Pediatric
Epistaxis. Otolaryngol Head Neck Surg. 2018 Oct;159(4):712-716. doi:
10.1177/0194599818785898. Epub 2018 Jul 10. PMID: 29986629.
• Send T, Bertlich M, Eichhorn KW, Ganschow R, Schafigh D, Horlbeck F,
Bootz F, Jakob M. Etiology, Management, and Outcome of Pediatric
Epistaxis. Pediatr Emerg Care. 2021 Sep 1;37(9):466-470. doi:
10.1097/PEC.0000000000001698. PMID: 30624421.
Editor's Notes
Medical Management: direct nasal compression, intranasal medications
Procedural Control: cautery and/or nasal packing
Data is shown as number of patients and the parentheses is percentage within the cohort.
Of 809 patients, 51 (6.3%) of patients received procedures.
The mean age of the entire cohort was 9.3 and the median was 8.6 years.
Patients who underwent procedures were older at 12 years old compared to 9 years old average, and median age of 13 compared to 8 years old.
55% of the total cohort was male and 45% was female. There was no difference between the medically managed and procedural group.
There was also no difference in race between the two groups.
While looking at epistaxis characteristics, we see that overall, 63% of children had nosebleeds under 30 minutes, and 34% had nosebleeds of 30+ minutes. 3% of children had missing data. Children who received procedures were more likely to have nosebleeds lasting 30 minutes or longer (80% compared to 31%). This was statistically significant.
59% of children presented with nosebleed of one nare, while 28% had nosebleed of both nares. 13% children had missing data. There was no statistically significant difference between the two groups.
Overall, 21% of children had no risk factor for nosebleed.
32% had a concurrent URI, 25% had direct trauma to the nose, and 18% had history of allergic rhinitis. Anticoagulation medication was the least common risk factor at 1%.
When comparing the two treatment groups, children with direct trauma were more likely to receive medical management.
Children with bleeding disorder were more likely to receive procedures (25% vs 5%) and children with antiplatelet medication (20% vs. 1%), as well as children with a known oncological condition and anticoagulation medication.
Using this data, we can build a logistic regression model for risk factors that increases odds for procedures in the ED for epistaxis.
The area under the ROC curve for this model was 0.85 (0.79-0.91).
This table shows the treatments that the patients received.
Overall, 80% of patients did not require any medications or procedures in the ED.
Overall, 19% of patients received an intranasal medication, and 84% of patients who received a procedure received an intranasal medication first prior to the procedure.
Vasoconstrictor were the most common intranasal medications used at 17% of the time. Children who received procedures were also associated with vasoconstrictor and antifibrinolytic medication use.
Within the procedural management group, 49% of patients received packing and 16% required both cauterization and packing.
Overall 3% of children received transfusion, and 16% of children who received procedures received a transfusion. There was statistical significance in platelet and blood & platelet transfusion between the two groups.
25% of children who received procedures were admitted compared to 3% of those who only received medical management. This was statistically significant.
Missing data implications are unknown
Low frequency of certain risk factors is due to smaller sample size, and there is great variability in the frequency of risk factors in the logistic regression model, thus this affects the overall fit of the model.