• 
• 

Three Part Question

In [adults without a bleeding disorder who have uncomplicated epistaxis] is [tranexamic acid] effective at [stopping bleeding and/or preventing rebleeding]?

Clinical Scenario

A 72-year-old man who is otherwise fit and well and on no regular medications presents to the emergency department (ED) with epistaxis that is ongoing despite appropriate first aid measures. No visible bleeding points can be seen on examination to allow cautery. After explaining treatment options to the patient, he states that he is very reluctant to have nasal packing because he once had it before and it was very uncomfortable and he then had to be admitted overnight, which he does not want. He is also concerned that after that admission he went home only to start bleeding again two days later. He wants to know whether there are any alternatives.

You have heard of people using tranexamic acid to stop epistaxis but you are not sure whether this was topical, oral or intravenous and you do not know whether there is any evidence to support this …

Search Strategy

A literature search was conducted using

Medline 1946 to May 2015 week 3 via NICE Healthcare Databases

EMBASE 1974 to present via NICE Healthcare Databases

CINAHL 1981 to present via NICE Healthcare Databases

Cochrane Library—“epistaxis”, “tranexamic acid”

Google Scholar—“epistaxis”, “tranexamic acid”

References of selected papers were reviewed for other relevant articles

Authors of the most relevant papers were contacted to see if any additional data was available but no reply has been received to date

Clinicaltrials.gov—“epistaxis”, “tranexamic acid”

The following search string was used: {exp epistaxis OR epistaxis.mp OR nosebleed$ OR “nose bleed*” OR rhinorrhag* OR “nasal haemorrhag*” OR “nasal hemorrhag*”} AND {exp “tranexamic acid” OR “tranexamic acid.mp” OR exp “antifibrinolytic agents” OR “antifibrinolytic agents.mp” OR cyklokapron}

Search Outcome

In total, 385 papers were found, of which 382 were excluded. The majority of the excluded papers were looking at the use of tranexamic acid for epistaxis during or after ear, nose and throat surgery or its use for patients with epistaxis who had a known bleeding disorder. These papers were not directly relevant to the aims of this BET so were excluded from the following table. However, many of the full papers were obtained and appraised separately to further understand the potential role for tranexamic acid in epistaxis. A Cochrane protocol looking at the use of tranexamic acid for epistaxis was published in 2003 and was updated with new authors in 2010. The review has yet to be published and was therefore also excluded. Other papers identified but excluded after full article review contained expert opinion regarding the use of tranexamic acid in epistaxis, which were mainly formed from the evidence in the papers included and extrapolated from use of the drug in patients with clotting disorders. Some of these are discussed in more detail later but were not considered to represent a sufficiently high level of evidence to include the table below

Relevant Paper(s)

Author, date and country	Patient group	Study type (level of evidence)	Outcomes	Key results	Study Weaknesses
Zahed R 2013 Iran	216 patients presenting with anterior epistaxis to a single ED.	Level 2 CEBM RCT. Cotton pledget soaked with 500mg in 5ml of tranexamic acid inserted into bleeding nostril vs anterior nasal packing with a cotton pledget soaked with epinephrine (1:100000) and 2% lidocaine for 10 minutes followed by packing with several cotton pledgets covered with tetracycline.	Percentage of patients whose bleeding stopped within 10 minutes.	71% of tranexamic acid group vs 31.2% of packing group (OR 2.28, 95% CI 1.68-3.09, P<.001).	Sampling method not reported. Not blinded to patients/attending physicians. No mention of first aid measures attempted to control bleeding prior to entering trial. Incomplete data presented to fully analyse baseline characteristics such as co-morbidities. Lack of information regarding the data collected such as actual time to termination of bleeding – only those less than 10 mins reported. Unclear when the time to cessation of bleeding was started in the anterior packing group – this group had an adrenaline/lignocaine pledget placed for 10 minutes prior to packing – not stated if this was counted in the treatment time. Gold standard/usual treatment is not currently multiple cotton pledget packing in UK EDs. Lignocaine and adrenaline only applied to packing group not both groups.
			Discharge from ED in less than 2hrs.	95.3% of tranexamic acid group vs 6.4% of packing group (OR 14.8, 95% CI 7.2 – 30.4, P<.001).
			Frequency of rebleeding in first 24 hrs and within 7 days.	Within 24 hrs: 4.7% in tranexamic acid group vs 12.8% in packing group (OR 0.36, 95% CI 0.14-0.98, P=0.34). Within 7 days: 2.8% in tranexamic acid group vs 11% in packing group (OR 0.26, 95% CI 0.07 – 0.88, P=.018).
			Patient satisfaction.	VAS score 8.5 (+/- 1.7) in tranexamic acid group vs 4.4 (+/- 1.8) in packing group (P<.001).
			Complications in the ED.	4.7% of tranexamic acid group vs 11% in packing group (OR 0.42, 95% CI 0.16-1.16, p=0.128).
Tibbelin A 1995 Sweden	68 Adult patients (>18yrs) with ongoing bleeding.	Level 2 CEBM RCT. Up to 15ml gel containing 10% tranexamic acid vs up to 15ml placebo gel administered into bleeding nostril via syringe.	Percentage of patients whose bleeding stopped within 30 minutes.	60% of tranexamic acid group vs 76% of placebo group (p=0.16).	Not clear where the study was conducted. Sampling method not reported. No sample size calculation. Not clear if all subjects accounted for as data presented only as percentages. No CONSORT diagram. No confidence intervals given. No mention of ethical approval. Method of randomization not reported. Baseline characteristics not comparable - Tranexamic acid group had significantly more moderate/heavy bleeds compared to more light bleeds in placebo group (p<0.01). Data in table not consistent with text (recorded in table as rebleeds in 30 and 8 days, in text mentioned as 10 days and 8 hours).
			Percentage of patients with rebleeding within 10 days.	44% of tranexamic acid group vs 66% of placebo group (not significant, p value not stated).
			Patient acceptance of treatment and adverse events.	No adverse events.
White A 1987 Scotland	96 adult patients (>18yrs) admitted to a single hospital with epistaxis. 89 included in final analysis.	Level 2 CEBM RCT. 1g of oral tranexamic acid TDS for 10 days vs placebo given within 1 hour of initial epistaxis control by cautery or nasal packing.	Percentage of patients with rebleeding within 3 weeks.	47% of patients taking tranexamic acid vs 57% of patients taking placebo (p > 0.50).	Study excluded patients taking warfarin, aspirin and dipyridamole plus anyone with renal impairment and anyone taking the oral contraceptive pill. No sample size estimates/power calculation performed. No CONSORT diagram. Unclear how sampling occurred. No recorded attempt to contact patients lost to follow-up. Per protocol analysis. No ethical approval recorded. No CIs given. Subjective measurements regarding classification of rebleeds as type 1/2/3 and categories defined by authors. Management of patients clearly different to current practice—including high number of patients receiving blood transfusions.
			Percentage of patients with rebleeding within 3 weeks but excluding first 24 hrs.	31% of patients taking tranexamic acid vs 45% of patients taking placebo (p > 0.10).
			Frequency of rebleeding.	40 episodes in tranexamic acid group vs 57 in placebo (x2 trends = 1.7, p=0.19).
			Frequency of type 1 rebleeds (mild).	13 episodes in tranexamic acid group vs 28 placebo (x2 trends = 4.9, p=0.03).
			Frequency of type 2 rebleeds (moderate).	13 episodes in tranexamic acid group vs 17 placebo (x2 trends = 0.2, p=0.6).
			Frequency of type 3 rebleeds (severe).	14 episodes in tranexamic acid group vs 12 in placebo (x2 trends = 0.01, p=0.9).

Comment(s)

None of the papers available are able to fully address the proposed aims of this BET, specifically to evaluate the efficacy of tranexamic acid in patients presenting to the ED with epistaxis. The heterogenous nature of the studies that are available make it difficult to combine the data. Each of the papers differs in regards to inclusion/exclusion criteria, treatments and outcome measures. Each of the studies appraised uses a different method of delivery of tranexamic acid, with two being topical usage and one oral. The paper by White and O’Reilly using oral tranexamic acid is designed to detect a reduction in recurrence of bleeding rather than the immediate control of epistaxis. However, prevention of recurrence of bleeding is relevant to the ED and our patient population. The randomised controlled trial (RCT) by Zahed et al is the most robust study and does indicate a benefit to the topical use of the injectable form of tranexamic acid; however, its multiple weaknesses as described above preclude it from being a definitive answer to the question. None of the studies find overall significant results, indicating a reduction in the recurrence of bleeding following use of tranexamic acid.

Editor Comment

CEBM, Centre for Evidence-Based Medicine; ED, emergency department; RCT, randomised controlled trial; VAS, visual analogue scale.

Clinical Bottom Line

The evidence is not sufficient to advise the routine use of tranexamic acid for patients attending ED with epistaxis. The evidence available is conflicting and there is a lack of sufficiently robust, well-designed trials on the topic to allow a recommendation to change current practice to include tranexamic acid. However, it is important to note that there is no evidence of tranexamic acid for epistaxis causing an increase in adverse events and the most robust study available currently shows a potential benefit. As such it is an area where more research would be beneficial. From the evidence available, it would seem that using the injectable form of tranexamic acid as a topical treatment would be the most likely to be beneficial and should be the starting point for future research. It may also be reasonable from the evidence currently available to offer topical tranexamic acid to patients such as the one in the clinical scenario above where the standard treatment of nasal packing is unacceptable to the patient.

References

1. Zahed R, Moharamzadeh P, AlizadehArasi S et al. A new and rapid method for epistaxis treatment using injectable form of tranexamic acid topically: a randomized controlled trial. American Journal of Emergency Medicine 2013; 31, 1389-1392.
2. Tibbelin A, Aust R, Bende M et al. Effect of local tranexamic acid gel in the treatment of epistaxis Journal for Oto-rhino-laryngology and its Related Specialities 1995; 57(4), 207-209
3. White A & O’Reilly BF Oral tranexamic acid in the management of epistaxis. Clin Otolaryngol Allied Sci 1988;13:11–16.