Is Continued Perioperative Antithrombotic Therapy Safe When Performing Open Tracheostomy?

Introduction

Tracheostomy is a commonly performed surgical procedure that restores the ability to breathe by creating an artificial airway through the neck and directly into the trachea. Common indications for tracheostomy include prolonged mechanical ventilation, inability to intubate, airway obstruction, need for pulmonary toilet, and airway protection. The total rate of complications from tracheostomy is approximately 15.8%, include bleeding, tracheal stenosis, tracheostomy tube plugging, and injury to the recurrent laryngeal nerve. ¹ The postoperative bleeding rate of tracheotomy ranges from 0.7% to 2.6%, making it a “high bleeding risk” procedure according to the criteria presented in a 2013 review in the New England Journal of Medicine, which defines clinical rates of bleeding of more than 1.5% as a high risk procedure. ²

The use of antithrombotic therapies (ATT) requires careful preoperative planning for patients who are undergoing tracheostomy. There are 2 categories of ATT based on their mechanism and intended purpose: antiplatelet therapies, which inhibit platelet activation and aggregation, and anticoagulation therapies, which target coagulation factors required for thrombin generation and fibrin formation. ³ There are no specific guidelines for ATT management at the time of an open tracheostomy, and the decision is often based on the surgeon’s experience. ⁴ Three critical factors to consider when determining whether or not to continue perioperative ATT use are: the patient’s inherent thromboembolic risk, the risk and potential consequences of procedure-related bleeding, and the timing of thromboembolic therapy interruption. ⁵ Surgeons have to weigh the risk of perioperative and postoperative bleeding if therapy is continued, the risk of thromboembolism if therapy is discontinued, and the benefits of performing the tracheostomy. Many forms of ATT can be reversed, although not commonly implemented, prior to tracheostomy. For example, vitamin K or fresh frozen plasma can be used for patients on warfarin and Idarucizumab, a monoclonal antibody treatment, can be administered for dabigatran reversal. ⁶

Unlike open tracheostomy, there has been substantial research performed on whether continued ATT use is safe during percutaneous tracheostomy. A 2020 study by Huang et al ⁷ with a cohort of 312 percutaneous tracheostomies, found that patients on ATT had a higher risk of bleeding (26.8%) than patients not on ATT (7.0%) during percutaneous dilational tracheostomy. They also looked at platelet counts in these patients and found that there was no relationship between bleeding and platelet count. ⁷ A more limited 2015 retrospective study performed by Pasin et al ⁴ with a cohort of 36 people observed that percutaneous dilational tracheostomy resulted in bleeding complications in 23% of patients on ATT.

While attention has been directed toward percutaneous tracheostomy, there is not enough data on the safety of perioperative ATT use during open tracheostomy. Hsueh et al ⁵ recommended that ATT be discontinued during open tracheostomy regardless of the patient’s risk for thromboembolism, however the broad nature of this recommendation may not be appropriate for every patient. Given that there is no reported significant difference in mortality, intraoperative hemorrhage, and postoperative hemorrhage between percutaneous and open tracheostomy, it is just as important to understand the best management of ATT during open tracheostomy as percutaneous tracheostomy. ⁸ As there are differences between these 2 procedures, including more control over the surgical field and hemostasis, we expect the risk of bleeding on ATT to be lower in open tracheostomy compared to percutaneous tracheostomy. In this study, we offer our institution’s experience with open tracheostomy patients. This investigation aims to evaluate how continuation or discontinuation of ATT during open tracheostomy influences perioperative and postoperative outcomes.

Materials and Methods

Following approval from the Institutional Review Board (IRB) at the Mayo Clinic, a retrospective review was performed on patients who underwent an open tracheostomy at a tertiary care medical center between January 2015 and December 2019. Patients were divided into 2 groups:

Variables were obtained for both groups including demographic data, smoking status, Body Mass Index (BMI), and comorbidities. The use of ATT before and during surgery was documented, including antiplatelet, anticoagulant, and direct anticoagulant use. Major outcome variables included perioperative bleeding, postoperative bleeding, and 30-day survival after tracheostomy, which are all strong indicators of safety. Secondary outcome variables included infection, ability to be decannulated, days to decannulation, and days to first tracheostomy change, which are all moderate indicators of safety.

Results

There were not enough patients in this study to perform the 3 and 4 group statistical analysis, therefore all results are reported for 2 groups: ATT versus non-ATT.

Demographics

A total of 217 tracheostomy patients were evaluated for this study, of which 148 (68.2%) were not on ATT and 69 (31.8%) were on ATT during surgery. Nine of 217 were revision tracheostomies. Basic demographic variables are included in Table 1. There was no significant difference observed on gender, race, BMI, or smoking status between the 2 groups. However, non-ATT patients were younger than ATT patients, with mean ages of 61.3 and 66.1 years respectively (P = .0178). All tracheostomies were open, and most were performed in the operating room (216, 99.5%) by the otolaryngology department (213, 98.2%). The 16 patients requiring awake tracheostomy were more likely to be non-ATT (15) than on ATT (1, P = .0226).

OPEN IN VIEWER

Table 1.

Comparing Demographics in Tracheostomy Patients Who were on Versus Off ATT During Their Surgery.

	Not on ATT n = 148 n (%)	On ATT n = 69 n (%)	P value
Average age, y	61.3	66.1	.0178 *
Gender			.7344
Male	93 (62.8)	45 (65.2)
Female	55 (37.2)	24 (34.8)
Race			.7440
Non-hispanic white	121 (81.8)	60 (87.0)
Black	6 (4.1)	3 (4.3)
Asian/Pacific Islander	4 (2.7)	2 (2.9)
Hispanic	5 (3.4)	2 (2.9)
American Indian	5 (3.4)	2 (2.9)
Other	4 (2.7)	0 (0.0)
Unknown	3 (2.0)	0 (0.0)
Average BMI (kg/m2)	27.6	28.0	.8272
Smoking			.2938
Never	54 (37.5)	32 (46.4)
Former	77 (53.5)	29 (42.0)
Current	13 (9.0)	8 (11.6)
Procedure type			.0226 *
Intubated tracheostomy	133 (89.9)	68 (98.6)
Awake tracheostomy	15 (10.1)	1 (1.4)

Abbreviation: BMI, body mass index.

*

Statistically significant.

Comorbidities

Comorbidities are included in Table 2. There was no significant difference in unweighted Charlson index score, severity weighted index, and age weighted index between the 2 groups. Patients on ATT were more likely to have a history of myocardial infarction, congestive heart failure, stroke/transient ischemic attack, and liver disease. ATT patients were significantly more likely to be thrombocytopenic than non-ATT patients. ATT patients were also more likely to have been intubated for a longer average number of days prior to tracheostomy.

OPEN IN VIEWER

Table 2.

Comparing Comorbidities in Tracheostomy Patients Who were on Versus Off ATT During Their Surgery.

	Not on ATT n = 148 n (%)	On ATT n = 69 n (%)	P value
Myocardial infarction	7 (4.7)	12 (17.4)	.0021 *
Congestive heart failure	40 (27.0)	31 (44.9)	.0089 *
Peripheral vascular disease	8 (5.4)	4 (5.8)	.9064
Stroke/Transient ischemic attack	15 (10.1)	21 (30.4)	.0002 *
Diabetes			.1423
Uncomplicated	31 (20.9)	23 (33.3)
End-organ damage	2 (1.4)	1 (1.4)
Liver disease	4 (2.7)	9 (13.0)	.0028 *
Chronic kidney disease	26 (17.6)	20 (29.0)	.0553
AIDS	1 (0.7)	0 (0.0)	.4937
Lymphoma	8 (5.4)	2 (2.9)	.4121
Leukemia	8 (5.4)	2 (2.9)	.4121
Dementia	1 (0.7)	2 (2.9)	.1916
COPD	11 (7.4)	6 (8.7)	.7471
Connective tissue disease	13 (8.8)	8 (11.6)	.5143
Hemiplegia	5 (3.4)	6 (8.7)	.0963
Peptic ulcer disease	3 (2.0)	1 (1.4)	.7683
Thrombocytopenia	31 (21.1)	28 (40.6)	.0027 *
Unweighted Charlson Index Score			.627
Mean (SD)	0.5 (1.0)	0.6 (1.2)
Median (IQR)	0 (0,1)	0 (0,1)
Severity of weighted sum of diseases			.504
Mean (SD)	1.0 (1.9)	0.9 (1.9)
Median (IQR)	0 (0,2)	0 (0,1)
Severity and age weighted sum of diseases			.415
Mean (SD)	2.8 (2.4)	3.0 (2.2)
Median (IQR)	2 (1,4)	2 (2,4)

Abbreviations: AIDS, acquired immune deficiency syndrome; COPD, chronic obstructive pulmonary disease.

*

Statistically significant.

Head and Neck Cancer

Overall, head and neck cancer patients were significantly less likely to be on ATT. Patients were less likely to be on ATT if they had prior neck radiation (5, 7.2%, P < .0001), prior neck dissection (1, 1.4%, P = .0069), prior tracheostomy (1, 1.4%, respectively), or presence of a solid head and neck tumor (20, 29.0%, P = .0001). Patients often had other concurrent procedures with a tracheostomy which can be found in Table 3. Having any concurrent procedure was significantly more common in non-ATT patients than ATT patients (59, 39.9%, and 7, 10.1%, respectively, P < .0001).

OPEN IN VIEWER

Table 3.

Comparing Surgical Details in Tracheostomy Patients Who were on Versus Off ATT During Their Surgery.

	Not on ATT n = 148 n (%)	On ATT n = 69 n (%)	P value
Reason			.0646
Cardiopulmonary failure	51 (34.5)	23 (33.3)
Obstructive malignancy	31 (20.9)	5 (7.2)
Progressive airway obstruction due to non-malignant cause	62 (41.9)	40 (58.0)
Conversion from cricothyroidotomy	2 (1.4)	0 (0.0)
Multiple	2 (1.4)	1 (1.4)
Average days intubated prior to tracheostomy, d	4.7	6.9	.0020 *
Any Concurrent Procedure	59 (39.9)	7 (10.1)	<.0001 *
Concurrent procedure
Reconstruction	26 (17.6)	4 (5.8)	.0193 *
Neck dissection	37 (25.0)	5 (7.2)	.0021 *
Tumor	31 (20.9)	3 (4.3)	.0017 *

*

Statistically significant.

Perioperative bleeding

Modeling of outcome variables is included in Table 4. There was no significant difference in incidence of perioperative bleeding based on ATT status (P = .983).

OPEN IN VIEWER

Table 4.

Modeling Primary Outcomes in Tracheostomy Patients Who were on Versus Off ATT During Their Surgery.

	Perioperative bleeding		Postoperative bleeding				30-day morality
	Univariate		Univariate		Multivariable		Univariate		Multivariable
Variable (Level)	OR (95% CI)	P value	OR (95% CI)	P value	OR (95% CI)	P value	OR (95% CI)	P value	OR (95% CI)	P value
Surgery blood thinner status (On ATT/Off ATT)	0.99 (0.36, 2.72)	.983	2.64 (0.92, 7.61)	.072	1.91 (0.65, 5.63)	.24	3.46 (1.49, 8.01)	.004 *	2.67 (1.07, 6.64)	.035 *
Any prior head and neck surgeries (Yes vs No)	1.44 (0.54, 3.85)	.464	0.16 (0.02, 1.22)	.077	0.27 (0.03, 2.16)	.215	0.53 (0.19, 1.48)	.227
History of MI (Yes vs No)	1.25 (0.27, 5.89)	.775	2.91 (0.74, 11.37)	.125			4.11 (1.41, 12)	.01 *	2.31 (0.7, 7.55)	.167
Intubated or awake tracheostomy (Awake vs Already Intubated)	4.13 (1.19, 14.4)	.026*	0 (0, )	.975			0.47 (0.06, 3.71)	.473
Days intubated prior to trach (One unit increase)	1 (0.92, 1.09)	.983	1.06 (0.99, 1.14)	.088	1.04 (0.97, 1.13)	.282	1.1 (1.04, 1.17)	.001 *	1.09 (1.02, 1.16)	.008 *
Any concurrent procedure (Yes vs No)	0.80 (0.28, 2.33)	.68	0.55 (0.15, 2.02)	.37			0.08 (0.01, 0.59)	.013 *

*

Statistically significant.

Awake tracheostomy patients had a significantly higher incidence of perioperative bleeding than patients who were intubated prior to tracheostomy (P = .026).

Discussion

This study has significant clinical relevance from both a patient’s and surgeon’s perspective. On average, over 100 000 tracheostomies are performed annually in the United States. ¹⁰ In a review of 5.4 million adults in 2014, 40.1% were using at least one anticoagulant. ¹¹ The frequency of tracheostomies, combined with the widespread use of ATT, makes it crucial to understand the safety of perioperative ATT during open tracheostomy. There can be disagreement between surgeons and intensive care unit (ICU) physicians regarding preoperative and postoperative care of patients. ¹² Having clearer guidelines on ATT management during open tracheostomy can help streamline these discussions and ultimately improve patient care.

Tracheostomies are considered a procedure with “high risk of bleeding.” However, the literature on the safety of perioperative ATT use during open tracheostomy is minimal. In our study, there was no significant difference in perioperative or postoperative bleeding based on ATT status, suggesting that continued ATT use during open tracheostomy is safe. This indicates that concerns about bleeding should not be used as the only reason for discontinuing ATT use during open tracheostomy.

Given the retrospective nature of this study, the decision to continue or discontinue ATT was based on a discussion between the surgeon and the intensivist without influence from the study. However, the general consensus was if the patient was on ATT due to a recent event, it was continued, and if they were on ATT for prophylaxis (eg, DVT prophylaxis) it was held.

The patients scheduled for surgery were mostly cancer patients, which does increase their risk of thrombotic events. However, they also tended to be non-emergent cases, meaning that there was more time for physicians to optimize their thrombotic therapy and stop ATT if possible. This is a generalization, as some surgical patients were urgent if their airway was compromised from tumor or radiation side effects.

It should also be noted that perioperative bleeding was increased in awake tracheostomies. Since awake tracheostomy patients were less likely to be on ATT, increased bleeding is likely not due to ATT status. Instead, it is likely due to the urgent or emergent nature of awake tracheostomies and the surgeon’s prioritization of speed over diligent hemostasis. Although there are no other published studies evaluating bleeding rates in awake tracheostomies, this finding is supported in the literature by a study that found the intraoperative hemorrhage rate to be 7.4% in percutaneous dilatational tracheotomy and 1.9% in open surgical tracheotomy ¹³ suggesting higher bleeding rates when there is not an opportunity for diligent hemostasis.

Although ATT patients were more likely to experience 30-day mortality, the increasing comorbidity of these patients and their causes of death indicate that ATT status alone was not the driving factor for 30-day mortality. This data was adjusted for history of myocardial infarction since there is a 5% re-infarction rate and high associated mortality rate for “patients undergoing non-cardiac surgery within 3 months of a myocardial infarction.” ¹⁴ This data was also adjusted for number of days intubated because individuals intubated for greater than 1 week are 1.42 times more likely to have late postoperative complications than those intubated for less than 1 week, however it is acknowledged that external factors such as timing of the consult and operating room availability may influence this number. ⁹ The significant correlation between 30-day mortality and increased number of days intubated prior to surgery supports this finding. Increased intubation time is associated with increased comorbidities or malignancy, and it stands to reason that their deaths were likely the result of several factors, not ATT status alone.

Further investigation into cause of death reveals that only 3/26 patients who experienced 30-day mortality had a cause of death related to bleeding. One patient was on ATT and died from severe hemorrhage at a site other than the tracheostomy. The other 2 patients were not on ATT. One died from severe hemorrhage at a site other than the tracheostomy. The other patient had a history of neck radiation and experienced bleeding at the tracheostomy site from a tracheoinnominate fistula 24 days after tracheostomy tube placement. Very few patients died from severe hemorrhage within 30 days after surgery, among whom only one was on ATT and they hemorrhaged from a site other than the tracheostomy. This allows us to conclude that ATT status during tracheostomy is not a major predictor of 30-day mortality.

There was no difference in the incidence of a major or minor thrombotic event within 30 days after surgery based on ATT status. This indicates that staying on ATT during surgery was effective in reducing thrombotic events for at-risk patients and posed no additional thrombotic risk after surgery.

Perioperative ATT use did not increase perioperative or postoperative bleeding, nor was it the single cause of any 30-day mortality, leading us to recommend that continued ATT use during open tracheostomy is safe. Further, since there is an increased risk of bleeding in patients on ATT undergoing a percutaneous tracheostomy, but no apparent risk in patients undergoing open tracheostomy, we recommend an open tracheostomy for patients who can’t safely stop their ATT.

Conclusion

Perioperative ATT use did not significantly impact the incidence of perioperative bleeding, postoperative bleeding, operative time, infection, ability to be decannulated, or days to decannulation. Patients on ATT were significantly more likely to experience 30-day mortality, however only one death was due to hemorrhage which was unrelated to tracheostomy in this group. Therefore, the authors suggest that continued perioperative ATT use is safe during open tracheostomy. However, given our discussion above, this decision should always be made by a clinician team.