Length of Stay among Patients Consulting for Spontaneous Epistaxis in the Emergency Department

Abstract

Spontaneous epistaxis is a common cause of emergency department’s (ED) visit. Considering ED’s overcrowding, it seems interesting to study risk factors of hospitalisation or long length of stay to increase triage’s quality. Thus, we performed a prospective analysis to assess average length of stay of patients with spontaneous epistaxis. Secondary objectives were to find putative risk factor of longer length of stay. From February 13th, 2023, to August 31st, 2023, a prospective observational study was performed in five EDs among patients who consult for spontaneous epistaxis. We collected the following time items – arrival, first contact with triage nurse, first medical contact, and discharged or admission time. We also collected sociodemographic, medical history, daily medications, and care in the ED. Among 106 patients, median length of stay in the ED was 144 min. No patient was discharged before 28 min and the longest duration was 625 min (10 h and 25 min. Half epistaxis was already dried up before their arrival in the ED. Main risks factors to long ED length of stay were the hour of admission between midnight and eight am, the need of a blood sample and a consult with an ORL. We did not find any impact of the location of care, hypertension, age, or blood thinners. Half of patient presenting in the ED for spontaneous epistaxis was discharged in 144 min. This duration is significantly prolonged during night shift, if blood was sampled, or if an ORL consult was needed.

Key Messages

Length of stay in the ED for a patient that consult for spontaneous anterior epistaxis is 2 h and 20 min.

Night and ORL consults increase this duration.

Hypertension, age, or blood thinners do not increase this duration.

Introduction

Spontaneous epistaxis, also known as nosebleed is a frequent disease [1]. Up to 60% of adults experiment at least one episode in a lifetime but only 6% seek medical attention [2]. Nose vascularization comes from a plexus directly branched from both internal (anterior ethmoidal) and external (sphenopalatine, greater palatine and superior labial) carotid arteries [3]. We estimate between 80 and 90% of epistaxis comes from the anterior nasal cavity, mainly from the anteroinferior septum in Little’s area, where the Kiesselbach plexus is found [4]. Although 94% of epistaxis cede spontaneously, it represented 0.5% of all emergency department (ED) visit – 100 000 per year in France [5]. Hospitalization is mandatory for 0.16% of all nosebleed [5, 6]. Two peaks of incidence have been described – the first around ten years old and the second between 70 and 79. It is also describes a seasonal impact with a higher incidence between December and February [6]. Older patients – up to 65 years old – consults more in the ED but also have more comorbidities such as hypertension [2]. They are more treated by anti-platelet and/or anticoagulant. Latest guidelines dated from 2019 by the American Academy of Otolaryngology-Head and Neck Surgery (ORL), mainly based on 2016 French ORL guidelines and an ‘‘audit’’ of epistaxis management from the United Kingdom [7–9]. It recommends first an identification to excluded other bleeding such as hematemesis of haemoptysis. Secondly, it recommends a prompt management as follow: after clearing blood-clots by blowing the nose or aspiration, ten minutes of compression located in the lower third of the nose [10, 11]. This step could be performed by the patient himself or its caregiver. If the bleeding persists, the patient needs to consult a physician, mainly in the ED. The clinician should perform anterior rhinoscopy to identify a source of bleeding after removal of any blood clot (if present), otherwise he should realise a nasal packing [7]. There is no mention of recommended time of surveillance after the bleeding has stopped. However, worldwide EDs are overcrowded [12, 13]. It seems interesting to identify patients that are susceptible to be hospitalized or stay longer in the ED. To our knowledge, no study has already analysed length of stay among patients consulting in the ED for spontaneous epistaxis. Thus, we performed a study to assess the length of stay of patients consulting for spontaneous epistaxis in the ED. We also want to find putative risk factors of increased length of stay.

Methods

Study Design

We performed a prospective observational study in five ED in Auvergne, France – the University Hospital at Clermont-Ferrand, Moulins-sur-Allier, Aurillac, Le-Puy-en-Velay, and Thiers. ED visits per year are respectively 60 000, 30 000 at Moulins and Aurillac, 42 000, and 15 000. Le Puy-en-Velay, Aurillac and Clermont-Ferrand’s ED could have ORL’s consult 24/7. In case of need, Moulins-sur-Allier and Thiers’ ED must call Clermont-Ferrand’s ORL, which is 1h30 from Moulins-sur-Allier and 40 min from Thiers. When a patient arrives in an ED, he must follow some steps which are presented in Fig. 1. The first one was administrative. He must present ID card, social security card and confirm date of birth, name, phone number and place where he lives. Next, he must wait in the waiting room before entering in the triage room. In this room, a nurse (with more than one year of experience [14]) asked him quick questions about medical history, daily treatments (such as blood thinners or antihypertensive medications), measure vital parameters (tension, cardiac frequency, oxygen saturation, capillary haemoglobin using Hemocue^®). At the end, according to the FRENCH scale [15], the patient is placed in an adapted place where he waited for a physician consult. Although a resident could care of a patient by himself, he must present the case to a physician before discharging or admit the patient in the hospital.

Fig. 1.

Different periods of management for patient with spontaneous epistaxis in an emergency department (ED). Non-medical time is the time between the entrance in the ED and the end of the time with the triage nurse. Medical time is from the first contact with the emergency doctor to the end of care. The total time of care is from the entrance to the discharge of the patient

Study Settings

We did not find any relevant and validated questionnaire on the bibliography. Thus, we – three EPs from different EDs – created a survey based on our experience and what we think relevant regarding our main objective. We used RedCap^®, with CHU Clermont-Ferrand’s license. It is composed of four parts. The first one was about time measurement, which were obtain directly from ED’s software. The second one was sociodemographic, location, medical history and daily medications, the third one about the actual episode of epistaxis (duration of bleeding, care performed…) and the last one about putative biology and ORL consult. It took eight minutes to perform this survey for each patient. Inclusion criteria was every patient consulting in the ED for epistaxis. Exclusion criteria were age under 18 years old, pregnancy, traumatic epistaxis, dementia, prisoners.

The main outcome is the total length of stay in the ED corresponding to the time between the creation of the administrative file to the discharge or admission of the patient. Secondary outcomes were to assess putative risk factors of longer or shorter length of stay, such as sociodemographic, location, daily medications, biology or ORL consult.

Data Analysis

The study aimed to detect a representative time of care from two groups of the same size with a Kaplan Meier analysis, so we estimated a population of at least 100 patients. Categorical data were expressed as number of subjects and associated percentages, and continuous data as mean ± standard deviation or median [25th; 75th percentile], according to the statistical distribution. Kaplan Meier approach was used as accurate approach to measure time lapses, and efficient way to discriminate subgroups. Statistical tests included t-test and χ² for continuous and categorical variables. Patients were excluded if main objective criteria were not fully completed or if length of stay was up to two standard deviation (2 σ). We analysed the factors that might influence the global time of care in the ED with a Cox Model. Statistical analysis was performed using the Stata^® software (version 15; StataCorp, College Station, Texas, USA). All tests were two-sided, with an alpha level set at 0.05. No correction for multiple testing was applied in the analysis of secondary outcomes or in subgroup analysis. Findings from these analyses were interpreted as exploratory.

Ethics

The study was approved by local Ethics Committee (IRB00013412, “CHU de Clermont Ferrand IRB #1”, IRB number 2022-CF064) with compliance to the French policy of individual data protection.

Results

Characteristics of Population

We included 106 patients between February 13th to August 31st, 2023, 31 in Clermont-Ferrand, 35 in Le-Puy-en-Velay, six in Aurillac, 28 in Moulins-sur-Allier and four in Thiers. One was excluded due to an age < 18-year-old and one due to incomplete data. Our population has a mean age of 72 years old ± 16. Sixty-two patients (59%) had an history of hypertension, 64 (61%) had anticoagulation and 37 (47%) antiplatelet therapy. Among 105 with completed time of admission, 70 (67%) consulted between eight am and six pm (8:00–18:00), 22 (21%) between six pm and midnight (18:00–00:00) and 13 (12%) between midnight and eight am (00:00–08:00). For 53 patients (50%), bleeding had already stopped before their arrival in the ED. Among 96 with completed information, 70 (73%) performed compression before ED consult. It was the first episode of epistaxis for 34 patients (35%) over 97 with completed information. Among 102 patients with completed data, 82 (80%) had a unilateral epistaxis. Vital parameters at triage were 149 ± 26 mmHg for systolic blood pressure, 85 ± 17 mmHg for diastolic blood pressure, 84 ± 17 pulses per minute and 96 ± 2% for oxygen saturation. Median duration of bleed before consulting was 90 min [Q1: 40 Q3: 180], i.e., one hour and 30 min. Only five (5%) were admitted and 39 (37%) had an ORL consult.

Main Objective

According to Kaplan-Meier survival curve (Figs. 2), 50% of patients were treated and discharged or admitted in 144 min i.e., two hours and 24 min between administrative entrance and discharge from ED. None was discharged before 28 min and the last patient to leave the ED was at ten hours and 25 min.

Fig. 2.

Length of stay of patients consulting the emergency department for epistaxis using Kaplan-Meier graph. min

Secondary Objective

We did not find any centre effect. Median lengths of stay were 154 [IQR = 106] in Clermont-Ferrand, 136 [115] in Le-Puy-en-Velay, 125 [107] in Aurillac, 148 [129] in Moulins and 73 [126] in Thiers. We also did not find any significant difference in each duration (before triage, before first medical contact).

Patients consulting after midnight stayed longer than those consulting between eight am and six pm, respectively 191 min [IQR = 141] versus 136.5 min [98], p = 0.03. Patients that benefit for a blood sample analysis stayed 41 min longer, 225 [100] versus 123.5 [110.5], p < 0.001. Bilateral epistaxis prolonged length of stay by 14 min, respectively 158.5 [194.5] versus 1435.5 [115], p = 0.036. Lastly, patients that need an ORL consult stayed minutes longer, 177 [141] versus 126.5 [110], p = 0.003.

We found a tendency of 25 min shorter length of stay between patient with a stopped bleeding before medical contact compared to active bleeding, respectively 131 [IQR = 127] versus 156 [105] minutes. Age as continuous variable or categorical, using group < 65, 65 to 75, 75 to 85 and up to 85 years old did not induces impact on length of stay. Interestingly, anticoagulation and/or antiplatelet therapies were not risk factors to stay longer, respectively 146 [93] and 148.5 [127.5]. Lastly, triage parameters were not significant risk factors to stay longer or shorter.

Only five patients were hospitalised for a median length of stay before admission of 250 min, 110 min longer than the other – all had an ORL consult before admission. Three were over 85 years old. Three came from Le-Puy-en-Velay, one from Clermont-Ferrand and one from Moulins.

Discussion

We found a duration of two hours and 24 min to care and discharge half of all patients consulting for spontaneous epistaxis, according to Kaplan-Meier method.

In 2013, French government realized a national study on more than 52,000 patients consulting in the ED. For one day, they studied all trajectories of all patients consulting in all French ED – from initial call to 15, SAMU (French 911) to decision of admission or discharging at the end of the care. Different steps, length of stay and orientation were studied. Among all patients discharged from ED after a consult, 48% were before two hours [16]. A similar observation was made in the 2022’FEDORU report, where 60% were discharged before four hours [17]. We found a 19-minute increased length of stay after midnight, that could be explained by the lower number of physicians in the ED during this period, compared to the day. Very interestingly, age and hypertension and anticoagulant-antiplatelet drugs are more frequent in our population – anticoagulation x 4.4, antiplatelet x3.7 and hypertension x2 [18–20]. However, we did not find any increased length of stay among those patients. This could be explained by a good management in the ED. Logically, waiting for a blood sample analysis lengthens the time of care in the ED (+ 56%). Furthermore, we did not study if performing a blood sample analysis changed the care. The main reason is because we did not expect that at the beginning of the study and so we did not ask ethics for this question. Only 2.9% of patients presenting in the ED for spontaneous epistaxis need transfusion [21]. We should really consider blood sample only in case of severe haemorrhagic, maybe if the shock index is over one (> 1) in triage or during surveillance [22]. In our population, only one patient had a positive shock index. Waiting for an ORL consult increased by 23% the length of stay in the ED. However, there was no significant change in the care. We could so ask if those consults are mandatory or at least relevant. Indeed, in some organisation, ORL must consult the patient in the ED to give them a new consult two or three days after to remove packing tissue. We could suggest creating daily consult available for ED without ORL need to decrease overcrowding. Only five patients were hospitalized in our sample, all after an ORL consult. Very interestingly, four patients came from EDs that cannot have access to a 24/7 consult. Mothership organisation like stroke or trauma system management seems logical and relevant for these patients [23].

However, our study has some limitations. First, it is an observational study, reducing the power and generalization of our results. We studied different EDs with different yearly admission levels and we did not have any difference of length of stay. Secondly, we did not follow patients after they were discharged, and we didn’t ask for enough specific information about the patient, to suggest putative reconsult within few hours after discharge, or within month. Thirdly we probably missed some patient. Indeed, it was not mandatory to perform the study. However, we were able to include enough patients to perform initial data analysis.

Conclusion

Between February 13th and August 31st, 2023, 106 patients consulted in five ED for a spontaneous epistaxis. From the moment that the patient arrive, half of them will be discharged in 144 min. None was discharged before 28 min and five were hospitalized. Risk factors to prolong ED length of stay were the hour of admission between midnight and eight am, the need of a blood sample analysis and a consult from ORL. We did not find any impact of the location of care, hypertension, age, or blood thinners.