Association of Early Steroid and Antibiotic Therapy With Airway Outcomes in Adult Epiglottitis: A 10‐Year Multicenter Retrospective Cohort Study

ABSTRACT

Background

Adult epiglottitis is an uncommon but potentially life‐threatening condition requiring rapid recognition and airway‐focused management, yet evidence regarding early medical therapy is limited. We evaluated temporal trends in adult epiglottitis and examined associations between early corticosteroid or antibiotic therapy and clinically important outcomes.

Methods

We conducted a retrospective cohort study of adults (≥ 18 years) presenting to U.S. emergency departments with acute epiglottitis between 2014 and 2024 using the TriNetX Research Network. Cases were identified using the ICD‐10 code J05.1; patients with preexisting tracheostomy were excluded. Early therapy was defined as systemic corticosteroid or parenteral antibiotic administration within 24 h of presentation. The primary outcome was endotracheal intubation; secondary outcomes included ICU admission, 30‐day ED recidivism, and 30‐day mortality. Propensity score matching (1:1) and multivariable Cox proportional hazards models were used to estimate risk ratios (RRs) and hazard ratios (HRs).

Results

The annual incidence proportion of adult epiglottitis increased from 0.002% in 2014 to 0.005% in 2024 (RR per year 1.09; 95% CI 1.08–1.10), with the steepest rise among adults aged 60–74 years. In matched cohorts, early corticosteroid therapy was associated with lower risks of intubation (RR 0.48; 95% CI 0.26–0.88) and ICU admission (RR 0.71; 95% CI 0.55–0.92). Early antibiotic therapy demonstrated similar associations for intubation (RR 0.54; 95% CI 0.34–0.85) and ICU admission (RR 0.72; 95% CI 0.60–0.86). In adjusted Cox models, early steroids (HR 0.33; 95% CI 0.25–0.44) and early antibiotics (HR 0.50; 95% CI 0.34–0.75) were independently associated with lower hazards of intubation. Older age, comorbidities, and hypoxia were strong predictors of airway compromise.

Conclusions

Adult epiglottitis incidence has increased over the past decade, particularly among older adults. Early corticosteroid and antibiotic therapy were independently associated with lower risks of intubation and ICU admission, supporting current clinical practice favoring timely medical therapy in adults with suspected epiglottitis.

1. Introduction

Acute epiglottitis is a rapidly progressive inflammatory condition of the epiglottis and supraglottic structures that can lead to sudden airway compromise and respiratory failure [1]. Although historically most common among children, widespread Haemophilus influenzae type b (Hib) vaccination has resulted in a dramatic epidemiologic shift, with adults now representing the majority of cases in high‐income countries [2, 3, 4]. Adult epiglottitis remains an uncommon but high‐stakes emergency department (ED) presentation, requiring timely recognition and coordinated airway management to prevent catastrophic deterioration.

Existing data on adult epiglottitis are limited by relatively small sample sizes, single‐center designs, and incomplete characterization of contemporary clinical practice [4, 5, 6, 7, 8, 9]. Prior studies suggest that comorbid conditions, advanced age, and physiologic derangements may predict airway intervention, but these findings have not been validated in large multicenter cohorts [1, 9, 10, 11, 12, 13]. Furthermore, while systemic corticosteroids and broad‐spectrum antibiotics are widely recommended, evidence supporting their early use is sparse and largely based on expert opinion. No large‐scale studies have directly evaluated the association between early medical therapy and invasive airway intervention or critical care admission in adults.

In parallel, recent reports from select regions have suggested a possible rise in the incidence of epiglottitis among adults, particularly older adults [2, 3, 10]. However, national‐ or network‐level epidemiologic trends remain poorly understood, and no prior work has examined temporal patterns across age strata using contemporary real‐world data.

To address these gaps, we conducted one of the largest multicenter studies of adult epiglottitis to date using the TriNetX Research Network, leveraging detailed electronic health record data from more than 100 U.S. healthcare systems. Our objectives were threefold: (1) to characterize temporal trends in the incidence of adult epiglottitis over the past decade; (2) to evaluate the association between early (< 24 h) corticosteroid or antibiotic therapy and clinically important outcomes, including endotracheal intubation and ICU admission; and (3) to identify demographic, comorbid, and physiologic predictors of airway compromise. These findings aim to inform evidence‐based management of adult epiglottitis in the ED and to guide future clinical and epidemiologic research.

2. Methods

2.1. Study Design and Data Source

We performed a large retrospective cohort study of adults presenting to the emergency department with acute epiglottitis using data from the TriNetX Research Network. This platform is a global, federated system that aggregates de‐identified electronic health record data from more than 100 participating healthcare organizations (HCOs), including academic medical centers, regional hospitals, and affiliated outpatient practices. The network provides real‐time access to clinical information for over 250 million patients across diverse care settings from 2014 onward. Data are sourced directly from the EHRs of participating institutions and include both structured elements—such as diagnoses, medications, and vital signs—and information extracted from narrative documentation through integrated natural language processing. TriNetX applies standardized data harmonization and quality assurance processes before data become queryable. To preserve privacy, only aggregated results and summary statistics are returned, and all patient‐level identifiers remain fully de‐identified throughout data processing and analysis.

2.2. Study Population, Variables, and Outcomes

We identified adults aged ≥ 18 years presenting to an emergency department (ED) with acute epiglottitis between January 1, 2014, and December 31, 2024. Epiglottitis cases were identified using the ICD‐10‐CM code J05.1 (Acute epiglottitis). The index date was defined as the earliest ED encounter with this diagnosis. To avoid misclassification of chronic or postoperative airway conditions, patients with an existing tracheostomy—identified by ICD‐10 diagnosis codes, CPT procedure codes, or device codes present prior to the index presentation—were excluded (Table S1).

Early systemic corticosteroid therapy was defined as receipt of any of the following agents within 24 h of ED presentation: hydrocortisone, prednisone, methylprednisolone, dexamethasone, or prednisolone. Early antibiotic therapy was defined as administration of at least one parenteral antibiotic active against common epiglottitis pathogens within the same 24‐h window, including ceftriaxone, ampicillin–sulbactam, piperacillin–tazobactam, or levofloxacin.

2.3. Statistical Analysis

Baseline characteristics extracted included age, sex, race, comorbidities (diabetes, ischemic heart disease, heart failure, chronic kidney disease, COPD, alcohol use disorder, and tobacco use), and initial vital signs (systolic blood pressure, heart rate, temperature, respiratory rate, and oxygen saturation). Comorbidities were defined using ICD‐10‐CM codes recorded within 1 year prior to presentation. Epinephrine administration at the index encounter was included as a marker of airway inflammation severity.

The primary outcome was endotracheal intubation within 30 days of the index encounter, identified using a validated combination of CPT, ICD‐10‐PCS, and SNOMED procedure codes. Secondary outcomes included ICU admission, 30‐day ED recidivism, and 30‐day all‐cause mortality. ICU admission was defined using billing and encounter‐level metadata indicating intensive care status. Mortality data derive from EHR documentation and linked death records available within TriNetX.

Steroid and antibiotic exposure cohorts were constructed separately; therefore, the number of eligible patients differed across exposure analyses because not all patients received each therapy within 24 h of presentation (Figure S1).

2.4. Propensity Score Matching

To reduce confounding, we performed 1:1 nearest‐neighbor propensity score matching without replacement for each exposure (early steroids and early antibiotics separately). Propensity scores were estimated using logistic regression including all demographic characteristics, comorbidities, and initial vital signs. Standardized differences < 0.1 were considered to indicate adequate covariate balance. The matched cohorts were used for risk‐ratio estimation of the primary and secondary outcomes.

No formal a priori sample size calculation was performed because this was a retrospective study using all eligible cases available within the TriNetX network during the prespecified study period. Propensity score matching was performed separately for the steroid and antibiotic exposure cohorts using 1:1 nearest‐neighbor matching without replacement. Covariate balance before and after matching was assessed using standardized mean differences (SMDs), with values < 0.10 considered to represent ideal balance. Because modest residual imbalance remained for selected physiologic variables after matching, multivariable Cox proportional hazards models included all baseline covariates used in matching, including respiratory rate and oxygen saturation, in addition to epinephrine administration.

2.5. Multivariable Regression Analyses

We constructed multivariable Cox proportional hazards models to evaluate the association between early therapy and time to intubation or ICU admission within 30 days. Multivariable Cox models included all baseline covariates used in matching, including respiratory rate and oxygen saturation, in addition to epinephrine administration. Hazard ratios (HRs) > 1 represent increased hazard of the outcome. Proportional hazards assumptions were evaluated using Schoenfeld residuals and visual inspection of log–log survival curves.

2.6. Epidemiological Trend Analysis

We evaluated temporal trends in adult epiglottitis incidence from January 1, 2014, to December 31, 2024, using the TriNetX Research Network. For each calendar year, we extracted the number of adult patients with an index ED encounter for acute epiglottitis (ICD‐10 code J05.1) and the total number of adult patients with an ED visit in the network during the same period. Incidence proportion was defined as the number of annual epiglottitis cases divided by the yearly adult ED population denominator, expressed as a percentage.

Age‐specific incidence was computed using the same methodology. To improve interpretability and mitigate sparsity in narrow age bands, age categories were collapsed into broader strata to improve interpretability and reduce sparsity in narrower age bands.

To assess temporal trends, we fit Poisson regression models with the yearly adult population as an offset term. The primary model estimated the association between calendar year and incidence proportion in the overall population. A second model included an interaction term between year and age category to evaluate whether temporal trends differed by age. Rate ratios (RRs) with 95% confidence intervals (CIs) were calculated by exponentiating the model coefficients. Model fit was assessed using deviance statistics, and statistical significance was defined as p < 0.05.

Risk ratios from matched cohorts were calculated directly from contingency tables with 95% confidence intervals (CIs). Cox regression results are presented as hazard ratios (HRs) with 95% CIs. All hypothesis tests were two‐sided with α = 0.05. Because of limited event counts in certain strata, formal sensitivity analyses evaluating effect modification by age or comorbidity were not performed. Such analyses may represent an important direction for future investigation. Trend modeling and figure generation were conducted in R (version 4.3.2), and all other analyses were performed using TriNetX's secure analytics environment. Variables were analyzed using available data as recorded within the TriNetX network. Completeness of variables used in the analyses was high, with all key covariates and outcomes demonstrating > 90% reportedness across participating institutions. Because data are derived from routine clinical documentation, missingness may vary across individual fields depending on local documentation practices. Because this study included a prespecified primary outcome and exploratory secondary analyses, we did not formally adjust for multiple comparisons. Secondary analyses should therefore be interpreted cautiously.

3. Results

Across the TriNetX network, the annual incidence proportion of adult epiglottitis increased steadily from 0.002% in 2014 to 0.005% in 2024 (Figure 1). In Poisson regression adjusted for yearly population size, calendar year was strongly associated with rising incidence (RR per year 1.09, 95% CI 1.08–1.10; p < 0.001), corresponding to an average 9% relative increase in incidence per year.

FIGURE 1.

Annual incidence proportion of adult epiglottitis overall and by age group, 2014–2024. Lines depict the yearly incidence proportion of emergency department presentations for acute epiglottitis in the TriNetX network, expressed as a percentage of all adult patients who presented to the emergency department in the network for that calendar year. The black line shows the overall incidence, and colored lines show incidence within collapsed age categories (18–29, 30–44, 45–59, 60–74, and ≥ 75 years). Incidence more than doubled over the study period, with the highest and steepest increases observed among adults aged 60–74 years.

When stratified by age, incidence increased across most demographic groups, although the magnitude of increase varied. Adults aged 18–29 years, the reference group, demonstrated a 7% annual rise in incidence (RR 1.07 per year; p < 0.001). Temporal trends were significantly steeper among adults aged 60–74 years, who experienced an additional 3.4% yearly increase compared with the reference group (interaction RR 1.03; p = 0.03), yielding an overall increase of approximately 10.7% per year. Trends among adults aged 30–44, 45–59, and ≥ 75 years did not differ significantly from the 18–29‐year group (interaction p > 0.14 for all). The steepest statistically significant increase was observed among adults aged 60–74 years.

3.1. Patient Characteristics and Early Steroid Therapy

A total of 4257 adults with acute epiglottitis received early (< 24 h) steroid therapy, and 238 did not. Before matching, patients who did not receive early steroids had a slightly lower prevalence of several comorbidities—including diabetes, COPD, chronic kidney disease, and heart failure—and generally higher oxygen saturation at presentation; however, most standardized differences were modest (Table 1). After matching (n = 163 per group), demographic variables, comorbidities, and presenting vital signs were generally balanced (all standardized differences < 0.12), although modest residual imbalance remained for selected physiologic variables. Epinephrine use at presentation remained infrequent and similar between groups after matching (both 9.2%).

TABLE 1.

Characteristics of patients who present to the emergency department with acute epiglottitis, stratified by whether they received early (< 24 h) steroid therapy. Vital signs were collected at the initial emergency department presentation. Categorical variables are presented as the total number of patients and valid percentages (%). Continuous variables are depicted as mean ±  standard deviation.

	Before matching			After matching
	Acute epiglottitis, early steroid therapy—no. (%)	Acute epiglottitis, no early steroid therapy—no. (%)	Std diff	Acute epiglottitis, early steroid therapy—no. (%)	Acute epiglottitis, no early steroid therapy—no. (%)	Std diff
	n = 4257	n = 238		n = 238	n = 238
Demographics
Age at index, years	50.4 ± 18.2	49.6 ± 21.5	0.0388	49.8 ± 18.5	49.6 ± 21.5	0.0119
Male sex	2769 (65.214)	103 (63.19)	0.0422	105 (64.417)	103 (63.19)	0.0255
White	2470 (58.172)	86 (52.761)	0.1090	94 (57.669)	86 (52.761)	0.0988
Black or African American	836 (19.689)	31 (19.018)	0.0170	27 (16.564)	31 (19.018)	0.0642
Comorbidities
Diabetes mellitus	905 (21.314)	29 (17.791)	0.0889	22 (13.497)	29 (17.791)	0.1184
Ischemic heart disease	608 (14.319)	15 (9.202)	0.1593	17 (10.429)	15 (9.202)	0.0412
Chronic kidney disease	176 (4.145)	12 (7.362)	0.1385	11 (6.748)	12 (7.362)	0.0240
COPD	395 (9.303)	10 (6.135)	0.1189	10 (6.135)	10 (6.135)	< 0.0001
Heart failure	353 (8.314)	10 (6.135)	0.0842	10 (6.135)	10 (6.135)	< 0.0001
Alcohol use disorder	257 (6.053)	10 (6.135)	0.0034	10 (6.135)	10 (6.135)	< 0.0001
Tobacco use	412 (9.703)	12 (7.362)	0.0839	10 (6.135)	12 (7.362)	0.0489
Presenting vitals
Systolic BP (mmHg)	128 ± 21.8	127 ± 19.6	0.0102	124 ± 20.7	127 ± 19.6	0.1831
Heart rate (bpm)	97.4 ± 5.81	96.4 ± 9.72	0.1248	97.3 ± 7.21	96.4 ± 9.72	0.1127
Temperature (°F)	83.7 ± 18.2	82.7 ± 18.5	0.0508	86.3 ± 21.6	82.7 ± 18.5	0.1754
Respiratory rate (breaths/min)	17.4 ± 3.51	16.8 ± 3.6	0.1576	18.0 ± 3.64	16.8 ± 3.6	0.3284
Oxygen saturation (%)	87.6 ± 19.9	94.2 ± 8.1	0.4357	85.1 ± 22.0	94.2 ± 8.1	0.5518
Medications at presentation
Epinephrine	720 (16.957)	15 (9.202)	0.2315	15 (9.202)	15 (9.202)	< 0.0001

Among matched patients, early steroid therapy was associated with significantly lower risks of intubation and ICU admission (Table 2). Intubation occurred in 8.6% of patients receiving early steroids compared with 17.8% of those who did not (RR 0.48, 95% CI 0.26–0.88; p = 0.017). Similarly, ICU admission was less frequent among patients treated with early steroids (35.6% vs. 50.9%; RR 0.71, 95% CI 0.55–0.92; p = 0.009). Thirty‐day ED recidivism and mortality were low in both groups and did not differ significantly.

TABLE 2.

Outcomes of patients who present to the emergency department with acute epiglottitis after propensity score matching, stratified by early (< 24 h) steroid administration. The primary and secondary outcomes were 30 days from initial presentation to the emergency department. Categorical variables are presented as the total number of patients and valid percentages (%).

Outcome	Acute epiglottitis, early steroid therapy—no. (%)	Acute epiglottitis, no early steroid therapy—no. (%)	RR	95% CI	p
	n = 163	n = 163
Intubation	14 (8.6)	29 (17.8)	0.48	0.26–0.88	0.017
ICU admission	58 (35.6)	83 (50.9)	0.71	0.55–0.92	0.009
30‐day ED recidivism	13 (7.9)	18 (11.0)	0.72	0.37–1.42	> 0.2
30‐day mortality	3 (1.8)	1 (0.6)	—	—	—

3.2. Patient Characteristics and Early Antibiotic Therapy

A total of 4041 patients received early antibiotics, and 454 did not. Before matching, the early antibiotic group was broadly similar demographically but had slightly higher rates of comorbidities such as COPD and heart failure, and lower oxygen saturation at presentation, suggesting greater initial illness severity (Table 3). After matching (n = 272 per group), groups were well balanced across demographic and clinical variables (standardized differences < 0.18 for all). Epinephrine administration remained similar between groups after matching (13.2% vs. 12.5%).

TABLE 3.

Characteristics of patients who present to the emergency department with acute epiglottitis, stratified by whether they received early (< 24 h) antibiotic therapy. Vital signs were collected at the initial emergency department presentation. Categorical variables are presented as the total number of patients and valid percentages (%). Continuous variables are depicted as mean ± standard deviation.

	Before matching			After matching
	Acute epiglottitis, early antibiotic therapy—no. (%)	Acute epiglottitis, no early antibiotic therapy—no. (%)	Std diff	Acute epiglottitis, early antibiotic therapy—no. (%)	Acute epiglottitis, no early antibiotic therapy—no. (%)	Std diff
	n = 4041	n = 454		n = 272	n = 272
Demographics
Age at index, years	49.9 ± 18.5	52.4 ± 21.1	0.13	51.4 ± 17.9	51.8 ± 20.4	0.02
Male sex	2385 (59.0)	263 (58.0)	0.02	158 (58.1)	154 (56.6)	0.03
White	2586 (64.0)	295 (65.0)	0.03	180 (66.2)	174 (64.0)	0.05
Black or African American	808 (20.0)	86 (19.0)	0.03	51 (18.8)	54 (19.9)	0.03
Comorbidities
Diabetes mellitus	929 (23.0)	95 (21.0)	0.05	55 (20.2)	58 (21.3)	0.03
Ischemic heart disease	594 (14.7)	63 (13.9)	0.02	42 (15.4)	38 (14.0)	0.04
Chronic kidney disease	376 (9.3)	40 (8.8)	0.02	23 (8.5)	24 (8.8)	0.01
COPD	395 (9.8)	35 (7.7)	0.07	20 (7.4)	22 (8.1)	0.03
Heart failure	336 (8.3)	27 (5.9)	0.09	19 (7.0)	17 (6.3)	0.03
Alcohol use disorder	198 (4.9)	33 (7.3)	0.10	17 (6.3)	18 (6.6)	0.01
Tobacco use	267 (6.6)	21 (4.6)	0.09	14 (5.1)	13 (4.8)	0.02
Presenting vitals
Systolic BP (mmHg)	125.7 ± 22.6	129.3 ± 27.4	0.14	126.5 ± 24.9	127.1 ± 25.2	0.02
Heart rate (bpm)	87.1 ± 21.2	86.4 ± 23.4	0.03	86.6 ± 17.2	86.9 ± 20.1	0.02
Temperature (°F)	97.6 ± 5.7	96.4 ± 8.9	0.16	98.1 ± 1.43	97.8 ± 3.21	0.12
Respiratory rate (breaths/min)	17.7 ± 5.6	18.9 ± 4.8	0.23	17.5 ± 4.2	18.2 ± 3.86	0.18
Oxygen saturation (%)	88.1 ± 21.3	91.7 ± 16.9	0.19	89.7 ± 22.3	90.9 ± 15.8	0.06
Medications at presentation
Epinephrine	661 (16.4)	61 (13.4)	0.08	36 (13.2)	34 (12.5)	0.02

In the matched antibiotic cohort, early antibiotic therapy was associated with lower rates of intubation and ICU admission (Table 4). Intubation occurred in 9.2% of patients receiving early antibiotics compared with 16.9% without early therapy (RR 0.54, 95% CI 0.34–0.85; p = 0.008). ICU admission occurred in 37.1% of early antibiotic patients versus 51.5% of controls (RR 0.72, 95% CI 0.60–0.86; p = 0.0004). Rates of 30‐day ED recidivism and mortality were similar between groups and overall low.

TABLE 4.

Outcomes of patients who present to the emergency department with acute epiglottitis after propensity score matching, stratified by early (< 24 h) antibiotic administration. The primary and secondary outcomes were 30 days from initial presentation to the emergency department. Categorical variables are presented as the total number of patients and valid percentages (%).

Outcome	Acute epiglottitis, early antibiotic therapy—no. (%)	Acute epiglottitis, no early antibiotic therapy—no. (%)	RR	95% CI	p
	n = 272	n = 272
Intubation	25 (9.2)	46 (16.9)	0.54	0.34–0.85	0.008
ICU admission	101 (37.1)	140 (51.5)	0.72	0.60–0.86	0.0004
30‐day ED recidivism	31 (11.4)	29 (10.7)	1.07	0.66–1.72	0.78
30‐day mortality	3 (1.1)	2 (0.7)	—	—	—

3.3. Multivariable Analysis

In multivariable Cox proportional hazards models adjusting for demographics, comorbidities, vital signs, and epinephrine use, both early steroid and early antibiotic therapy were independently associated with substantially lower hazards of intubation and ICU admission (Figure 2).

FIGURE 2.

Forest plots displaying adjusted hazard ratios (HRs) and 95% confidence intervals for the association between early (< 24 h) steroid or antibiotic therapy and two clinical outcomes—30‐day endotracheal intubation (top panels) and 30‐day ICU admission (bottom panels)—among adults presenting to the emergency department with acute epiglottitis. Each panel represents a separate multivariable Cox proportional hazards model adjusting for demographic characteristics, comorbidities, presenting vital signs, and epinephrine administration. Triangles denote the primary exposure of interest (early steroid or early antibiotic therapy), and circles represent other covariates included in the model. The vertical reference line indicates HR = 1.

For intubation, early steroid therapy was associated with a lower hazard of intubation within 30 days (HR 0.33, 95% CI 0.25–0.44; p < 0.0001), while early antibiotic therapy was also associated with a lower hazard (HR 0.50, 95% CI 0.34–0.75; p = 0.0006; Tables S2 and S3). Increasing age, diabetes, heart failure, chronic kidney disease, and alcohol use disorder were each independently associated with higher hazards of intubation, whereas lower systolic blood pressure, lower temperature, higher respiratory rate, lower oxygen saturation, and the administration of epinephrine were all strong physiologic predictors of intubation.

For ICU admission, early steroid therapy was associated with a significantly lower hazard of ICU admission (HR 0.50, 95% CI 0.27–0.73; p < 0.0001), as was early antibiotic therapy (HR 0.51, 95% CI 0.35–0.73; p = 0.0003; Tables S4 and S5). Similar to the intubation models, older age, diabetes, heart failure, chronic kidney disease, and alcohol use disorder were independently associated with increased risk of ICU admission, while lower systolic blood pressure, higher respiratory rate, and lower oxygen saturation were strong physiologic predictors of ICU utilization. Epinephrine administration consistently exhibited the largest effect size across all models, with a fourfold to fivefold increase in hazard, reinforcing its role as a marker of severe airway compromise at presentation.

4. Discussion

In this large, multicenter cohort study leveraging real‐world data from the TriNetX network, we provide what is, to our knowledge, the most comprehensive epidemiologic and outcomes‐focused evaluation of adult epiglottitis conducted to date. Prior investigations have largely been limited to single‐center experiences, administrative datasets with limited clinical detail, or small retrospective cohorts in which sample size constrained the ability to evaluate treatment‐related outcomes [5, 6, 8, 9, 11, 13]. Our analysis, which included more than 4000 adults and enabled propensity‐matched and fully adjusted multivariable modeling, offers new insights into both the changing epidemiology of adult epiglottitis and the association of early medical therapy with airway and critical care outcomes.

The finding that the incidence proportion of adult epiglottitis has nearly tripled over the past decade adds important context to a condition historically considered rare in the post– Haemophilus influenzae type b vaccination era. While population‐level increases have been reported in select national datasets, few studies have evaluated age‐specific trends [2, 3, 10]. Our analysis demonstrates that incidence has risen across most adult age groups, with the steepest increases observed among adults aged 60–74 years. These data suggest a shifting demographic profile of adult epiglottitis, with older adults now representing a disproportionately growing share of cases. Several mechanisms may contribute to the increasing incidence observed among older adults. Population aging and a rising burden of chronic disease may increase susceptibility to airway infections or inflammatory conditions affecting the supraglottic structures. Age‐related immunologic changes and higher rates of comorbid conditions such as diabetes or chronic kidney disease may also increase vulnerability to severe presentations. Additionally, improved recognition and diagnostic coding of epiglottitis in adults may partially contribute to the observed rise.

Our study also provides the largest assessment to date of early medical therapy—including systemic corticosteroids and parenteral antibiotics—and their associations with clinically meaningful outcomes. Existing literature on the role of steroids in epiglottitis is inconsistent, with small observational series reporting mixed results, and no randomized trials available [14, 15, 16, 17, 18]. Similarly, although antibiotic therapy is universally recommended, prior studies have not examined whether earlier administration confers measurable benefit. Using propensity‐matched cohorts and multivariable Cox regression adjusting for demographics, comorbidities, physiologic markers, and epinephrine use, we found that both early steroid and early antibiotic therapy were independently associated with significantly lower hazards of intubation and ICU admission. The magnitude of these associations—approximately 50%–70% relative reductions—remained robust across analytic approaches. These findings provide new observational evidence consistent with current practice favoring timely medical therapy in adults with suspected epiglottitis, complementing existing guideline recommendations that have been based largely on expert consensus.

Additionally, our analyses identified several clinical and physiologic predictors of airway compromise that align with prior reports. Older age, diabetes, heart failure, chronic kidney disease, and alcohol use disorder were consistently associated with worse outcomes, confirming risk factors described in other studies [12, 19]. Vital sign abnormalities, such as lower systolic blood pressure, tachypnea, and hypoxia, emerged as strong physiologic markers of impending airway deterioration. Epinephrine administration—likely reflecting clinician recognition of severe airway inflammation or stridor—was among the strongest predictors of both intubation and ICU use, underscoring its role as a surrogate indicator of severity rather than a therapeutic signal.

Our study has important clinical implications. Given the rapid progression that can characterize adult epiglottitis, early identification of high‐risk patients remains paramount. The associations observed between early medical therapy and lower risks of intubation and ICU admission suggest that timely diagnostic and treatment pathways may be beneficial in this population. At the same time, the rising incidence among older adults highlights the need for heightened vigilance in emergency and acute care settings as the population continues to age. From a clinical perspective, these findings support current practice emphasizing early recognition, close airway monitoring, and prompt initiation of empiric antimicrobial therapy and corticosteroids in adults with suspected epiglottitis while definitive evaluation is pursued.

This study has several important limitations. First, as a retrospective analysis of electronic health record data from the TriNetX network, our findings are subject to residual confounding and potential misclassification despite extensive adjustment and propensity matching. TriNetX does not provide granular clinical details such as symptom onset, airway exam findings, radiographic interpretations, or precise indications for intubation, which limits our ability to differentiate prophylactic from emergent airway management. Medication administration timing is captured within broad 24‐h windows, and we cannot assess dosing, specific steroid or antibiotic regimens, or adherence to full treatment courses. Although propensity score matching improved covariate balance, unmeasured confounders—particularly severity‐of‐illness markers not captured in structured fields—may still bias estimates toward or away from benefit. Epiglottitis is also a heterogeneous condition, and ICD codes cannot reliably distinguish infectious from noninfectious etiologies, nor differentiate supraglottitis, vallecular inflammation, or epiglottic abscess. Because TriNetX primarily represents large academic and affiliated community hospitals, the cohort may not generalize to smaller or resource‐limited settings. Moreover, the observational nature of the study precludes establishing causality; the associations observed between early therapy and improved outcomes should be interpreted as hypothesis‐generating and require validation in prospective or interventional studies.

Although intubation and ICU admission are relatively objective outcomes, ascertainment of 30‐day ED recidivism and mortality may be more susceptible to coding inconsistency or incomplete capture across institutions. In addition, follow‐up outcomes were limited to events recorded within participating health systems and linked records; therefore, encounters occurring outside the network may not have been captured. We were also unable to assess longer‐term outcomes such as recurrent epiglottitis, delayed airway complications, dysphagia, or quality of life. Finally, microbiologic data were not available in sufficient detail to evaluate pathogen‐specific epidemiology or determine whether treatment associations differed by infectious etiology.

Future work should explore the underlying drivers of rising incidence, evaluate microbial and immunologic contributors, and determine whether specific steroid regimens or antibiotic combinations confer differential benefit. Prospective studies—even if pragmatic or registry‐based—may help clarify causal relationships and refine treatment algorithms.

Overall, this study provides updated epidemiologic data on adult epiglottitis and adds observational evidence that early corticosteroid and antibiotic therapy are associated with lower risk of airway compromise and ICU utilization. These findings help fill a longstanding gap in the literature and may inform both clinical decision‐making and guideline development for this rapidly progressive and potentially life‐threatening condition.

Author Contributions

Joseph O'Brien conceived and designed the study, performed data analysis and interpretation, drafted the manuscript, provided statistical expertise, and critically revised the manuscript for important intellectual content. Dominic R. Muccio contributed to the study design, interpretation of the data, and critical revision of the manuscript for important intellectual content. Jon W. Schrock contributed to the study design, interpretation of the data, and critical revision of the manuscript for important intellectual content. All authors approved the final manuscript as submitted.

Funding

The authors have nothing to report.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Table S1: Query codes and definitions used to identify the study cohort, exposures, and outcomes.

Figure S1: Study flow diagram for exposure‐specific cohort construction. Flow diagram showing derivation of the adult epiglottitis cohorts used for the propensity score–matched analyses. Panel A depicts the corticosteroid cohort, including adults presenting to the emergency department with acute epiglottitis who did and did not receive early systemic corticosteroid therapy within 24 h of presentation. Panel B depicts the antibiotic cohort, including adults who did and did not receive early parenteral antibiotic therapy within 24 h of presentation. Exposure cohorts were constructed separately; therefore, the number of eligible patients differed between the steroid and antibiotic analyses. Outcomes were compared after 1:1 propensity score matching within each exposure‐specific cohort.

Table S2: Multivariable Cox proportional hazards regression evaluating the association between early (< 24 h) steroid administration and 30‐day intubation among adults presenting to the emergency department with acute epiglottitis. The model adjusts for demographic characteristics, comorbidities, presenting vital signs, and receipt of epinephrine. Hazard ratios > 1 indicate increased hazard of intubation.

Table S3: Multivariable Cox proportional hazards regression evaluating the association between early (< 24 h) antibiotic administration and 30‐day intubation among adults presenting to the emergency department with acute epiglottitis. The model adjusts for demographic characteristics, comorbidities, initial vital signs, and epinephrine administration at presentation. Hazard ratios > 1 indicate increased hazard of intubation.

Table S4: Multivariable Cox proportional hazards regression evaluating the association between early (< 24 h) steroid therapy and 30‐day ICU admission among adults presenting to the emergency department with acute epiglottitis. The model adjusts for demographic characteristics, comorbidities, presenting vital signs, and epinephrine administration. Hazard ratios > 1 indicate an increased hazard of ICU admission.

Table S5: Multivariable Cox proportional hazards regression evaluating the association between early (< 24 h) antibiotic therapy and 30‐day ICU admission among adults presenting to the emergency department with acute epiglottitis. The model adjusts for demographic characteristics, comorbidities, presenting vital signs, and epinephrine administration. Hazard ratios > 1 indicate increased hazard of ICU admission.

O'Brien J., Muccio D. R., and Schrock J. W., “Association of Early Steroid and Antibiotic Therapy With Airway Outcomes in Adult Epiglottitis: A 10‐Year Multicenter Retrospective Cohort Study,” Academic Emergency Medicine 33, no. 4 (2026): e70285, 10.1111/acem.70285.

Data Availability Statement

The data that support the findings of this study are available from TriNetX. Restrictions apply to the availability of these data, which were used under license for this study. Data are available from the author(s) with the permission of TriNetX.