Tracheotomy Outcomes in the Medical Intensive Care Unit

Abstract

This is a retrospective chart review of 161 MICU patients who underwent tracheotomy by the Otolaryngology service at the University of Missouri Hospital from April 2015 through March 2018. The objectives of this study were to describe long term outcomes of patients who underwent tracheotomy, and identify any clinical variables associated with these outcomes. One-year mortality was 37% (50/161), and on multivariate analysis only age (p=0.001) was associated with mortality.

Introduction

Mechanical ventilation for patients with acute respiratory failure in the intensive care unit (ICU) is delivered via endotracheal tube (ETT) or tracheostomy tube. In addition to pulmonary toilet, airway obstruction, and the need for secure surgical airway, indications for tracheotomy include anticipated prolonged ventilation or ventilatory problems. Patients requiring prolonged mechanical ventilation commonly are candidates for tracheostomy, a safe and routine surgical procedure, especially when performed on an elective basis.¹ While the surgical procedure itself is considered safe, inpatient mortality rates remain high after tracheostomy, on the order of 28 to 39%.^2–5 Other groups have shown that post-ICU mortality was increased in patients with a tracheostomy, although this may be due to confounding comorbidities that prevented decannulation and increased those patients’ risk for death at baseline.^4,5

Although data regarding the impact of tracheostomy on mortality remains conflicted, the idea that tracheostomy improves patient quality of life is well-supported. Compared to intubation, tracheostomy is associated with increased patient comfort, increased ease of tube replacement and suctioning, and the ability to transfer the patient from the ICU. Patient speech, mobility, and swallowing are enhanced following tracheostomy.⁶ Tracheostomy is associated with a decrease in sedation time (from seven to one hour per day) with no increase in agitation.⁷

Tracheotomy is the most commonly performed operation in critically-ill patients, and approximately two-thirds of all tracheostomies performed are on critically-ill patients requiring mechanical ventilation in the ICU.^8–10 Describing outcomes such as inpatient decannulation, discharge disposition, and one-year mortality as well as identifying any clinical variables associated with these outcomes will allow for better preoperative counseling when discussing these difficult decisions. Additionally, as outcomes are becoming more intensely scrutinized by third-party payers, published data describing the outcomes of these patients is important. The objectives of this study are to report inpatient mortality and one-year mortality after tracheotomy in the medical intensive care unit (MICU) patient, and to identify any clinical features associated with one-year mortality. Additionally, outcomes regarding ventilator liberation, tracheotomy decannulation, and disposition are reported.

Methods

This study was approved by the University of Missouri Institutional Review Board. This is a case series with chart review of all MICU patients who underwent tracheotomy by the Otolaryngology service at the University of Missouri Hospital from April 2015 through March 2018. The operative notes, clinical notes, laboratory data, and vital signs were reviewed. Demographic, clinical, and follow-up data were collected. Percutaneous versus open technique and BMI were recorded. The world health organization has defined three subclasses of obesity: obese class 1 (BMI 30.00 – 34.99 kg/ m²), obese class 2 (BMI 35.00 – 39.99 kg/ m²), and obese class 3 (BMI ≥ 40.00 kg/ m²).¹¹ For study purposes patients were classified as not obese (BMI < 30.00 kg/ m²), class 1 or 2 obesity (BMI 30.00 – 39.99 kg/ m²), and class 3 obesity (BMI ≥ 40.00 kg/ m²). The presence or absence of the following medical comorbidities were recorded: type 2 diabetes mellitus (T2DM), hypertension, heart failure, coronary artery disease, pulmonary hypertension, chronic obstructive pulmonary disease, chronic kidney disease, and obstructive sleep apnea. The Acute Physiology and Chronic Health Evaluation II (APACHE II) score was calculated for each patient.¹² The study outcome measures were time from tracheotomy to liberation from ventilator, capping tracheotomy, decannulation, MICU discharge, hospital discharge, duration of hospital admission, duration of MICU admission, discharge disposition, and one-year mortality.

Statistical Analysis

Mean age and APACHE II score were compared across BMI groups using analysis of variance. Sex, surgical technique, comorbidities, inpatient decannulation, disposition, and one-year mortality were compared across BMI groups using the Pearson Chi-Square test. Time from tracheotomy to liberation from ventilator, capping tracheotomy, decannulation, MICU discharge, hospital discharge, duration of hospital admission, and duration of MICU admission were compared across BMI groups using analysis of variance. Multivariable logistic regression was used to analyze the association of age, tracheotomy technique, BMI category, APACHE II score, and gender with one-year mortality. The statistical analysis was completed with IBM SPSS Statistics Version 25 (SPSS, Inc, and IBM Company, Armonk, New York, USA).

Results

A total of 161 MICU patients underwent tracheotomy during the study period. The mean age of patients included in the study was 58 years (range: 20 – 83 years). The study population was 58% (93/161) female and 42% (68/161) male. The mean BMI of patients included in the study was 35.2 kg/ m² (range: 15 – 89 kg/ m²). Forty-one percent (66/161) of patients were not obese, 28.6% (46/161) were class 1 or 2 obese, and 30.4% (49/161) were class 3 obese (Table 1). Twelve percent (19/161) were done using a percutaneous technique, and those patients who were not obese were significantly more likely to undergo a tracheotomy using this technique (p=0.037). Sixty-nine percent (34/49) of patients with class 3 obesity had T2DM, while 26% (17/66) of non-obese patients had T2DM (p< 0.001). Twenty-nine percent (14/49) of class 3 obese patients had heart failure, while 9% (6/66) non-obese patients had heart failure (p=0.016). Thirty-nine percent (19/49) of class 3 obese patients had obstructive sleep apnea, which 3% (2/66) non-obese patients had obstructive sleep apnea (p< 0.001). Inpatient decannulation rates for non-obese, class 1 and 2 obese, and class three obese patients were 26% (17/66), 15% (7/46), and 24% (12/49); respectively (p=0.417). Discharge disposition to facility for non-obese, class 1 and 2 obese, and class three obese patients was 48% (32/66), 70% (32/46), and 55% (27/49); respectively (p=0.097). The overall inpatient mortality rate was 15% (24/161). The overall one-year mortality rate was 37% (59/161). One-year mortality for non-obese, class 1 and 2 obese, and class three obese patients was 41% (27/66), 35% (16/46), and 33% (16/49); respectively (p=0.631). Of the 59 patients who died during the first year after tracheotomy, the range in days from tracheotomy to mortality was 14 – 358 days, with an average of 98 days.

Table 1.

Comorbidities, Clinical features, and Clinical Outcomes by Body Mass Index Category

	BMI <30	BMI 30–40	BMI > 40	P value
N	66	46	49
Age	56.7	60.8	57.1	P=0.261
Sex
Female	33 (50)	26 (57)	34 (69)	P=0.112
Male	33 (50)	20 (43)	15 (31)
Technique
Open	56 (85)	38 (83)	48 (98)	P=0.037*
Percutaneous	10 (15)	8 (17)	1 (2)
Comorbidities
T2DM	17 (26)	13 (28)	34 (69)	P<0.001*
Hypertension	36 (55)	32 (70)	35 (71)	P=0.114
Heart Failure	6 (9)	12 (26)	14 (29)	P=0.016*
CAD	18 (27)	19 (41)	16 (33)	P=0.298
PH	0 (0)	2 (4)	4 (8)	P=0.071
COPD	29 (44)	23 (50)	25 (51)	P=0.709
CKD	8 (12)	7 (15)	7 (78)	P=0.885
OSA	2 (3)	6 (13)	19 (39)	P<0.001*
APACHE II	16.7	16.9	16.9	P=0.990
Inpatient Decannulation	17 (26)	7 (15)	12 (24)	P=0.417
Disposition
Home	18 (27)	6 (13)	13 (27)	P=0.097
Facility	32 (48)	32 (70)	27 (55)
One Year Mortality	27 (41)	16 (35)	16 (33)	P=0.631

^*Indicates statistical significance.

Abbreviations: BMI, body mass index; T2DM, type 2 diabetes mellitus; CAD, coronary artery disease; PH, pulmonary hypertension; COPD, chronic obstructive pulmonary disease; CKD, chronic kidney disease; OSA, obstructive sleep apnea; APACHE II, Acute Physiology and Chronic Health Evaluation II.

Time from tracheotomy to ventilator liberation for non-obese, class 1 and 2 obese, and class three obese patients was 15.4 days, 16.0 days, and 13.9 days; respectively (p=0.842) (Table 2). Time from tracheotomy to decannulation for non-obese, class 1 and 2 obese, and class three obese patients was 35.5 days, 48.6 days, and 44.7 days; respectively (p=0.506). Time from tracheotomy to hospital discharge for non-obese, class 1 and 2 obese, and class three obese patients was 27.9 days, 30.3 days, and 30.0 days; respectively (p=0.835). On multivariate analysis, only age was significantly associated with one-year mortality (p=0.001), while sex (p=0.127), surgical technique (p=0.764), BMI category (p=0.504), and APACHE II score (p=0.560) were not significantly associated with one-year mortality (Table 3).

Table 2.

Time in Days for Inpatient Care Transition Milestones

	BMI <30	BMI 30–40	BMI >40	P value
Tracheotomy to ventilator liberation	15.4	16.0	13.9	P=0.842
Tracheotomy to capping	21.0	24.2	25.9	P=0.481
Tracheotomy to decannulation	35.5	48.6	44.67	P=0.506
Tracheotomy to MICU discharge	13.7	17.1	16.2	P=0.415
Tracheotomy to hospital discharge	27.9	30.3	30.0	P=0.835
Hospital admission duration	41.1	43.2	42.0	P=0.896
MICU admission duration	26.8	30.0	28.2	P=0.574

Table 3.

Multivariable Binary Logistic Regression

Variable	Multivariate Analysis: Hazard Ratio (95% CI)	P Value
Age	1.05 (1.02–1.08)	P=0.001*
Sex
Female	0.57 (0.28–1.71)	P=0.127
Male	Reference
Technique
Open	1.20 (0.37–3.85)	P=0.764
Percutaneous	Reference
BMI Category
<30	1.33 (0.58–3.08)	P=0.504
30–40	0.87 (0.35–2.17)	P=0.761
>40	Reference
APACHE II	1.02 (0.96–1.08)	P=0.560

^*Indicates statistical significance.

Abbreviations: BMI, body mass index; APACHE II, Acute Physiology and Chronic Health Evaluation II.

Discussion

Treating patients in the ICU necessitates a complex decision-making process for all of those involved, including medical staff, patients, and their family members. Patients’ intubated and sedated status, inability to effectively communicate or take oral nutrition, the need for mechanical ventilation, and the difficult decisions that need to be made about patient care all contribute to this process. Prolonged intubation can lead to further complications, such as laryngeal edema, subglottic stenosis, complications at the cuff site, tracheal bleeding, and injuries at the mouth or nose, making all of these factors mentioned intensified.¹³ Additionally, ETT intubation can result in laryngeal ulceration which can evolve into posterior glottic stenosis, a very challenging disease to correct.¹⁴

While tracheostomy is a reliable tool to assist with improvement in outcomes for patients by decreasing their risk of development of these complications from ETT intubation, the clinical requirement of tracheostomy alone can predict that these patients have a high risk of mortality and palliative care discussions should be utilized.

Our study examines the outcomes of critically-ill patients in the MICU undergoing tracheostomy, specifically identifying inpatient mortality and one-year mortality rate. We found the inpatient and one-year mortality rate to be 15% and 37%, respectively, which is similar to previous studies.^15,16 Other than age, there were no significant risk factors identified to help stratify mortality risk of critically-ill patients undergoing tracheostomy.

It has been found in a randomized trial that two-year all-cause mortality was 51% in patients requiring prolonged mechanical ventilation who underwent early tracheostomy (within four days) and 53.7% in patients requiring prolonged mechanical ventilation who underwent late tracheostomy (after 10 days if still indicated).¹⁷ In the late tracheostomy group, 53.7% of patients never underwent tracheostomy.¹⁷ Evidence supports that risk for mortality is high in MICU patients requiring prolonged mechanical ventilation, regardless of whether the patient undergoes tracheotomy.¹⁷ The inpatient and one-year mortality rates that we report in this current study can be discussed with the patient family during the preoperative discussion. As the rate of all-cause mortality in this patient population is high, a consideration of palliative care discussion with the patient and family is warranted, with a focus on patient goals of care and quality of life.

The mortality rate is high in all MICU patients requiring prolonged mechanical ventilation, and the mortality rates after tracheotomy in this patient population seen in this study and other published studies should not be interpreted as failure of the tracheotomy procedure. In fact, we would argue that the tracheotomy likely provided palliation to those patients who died. Studies have shown that critically-ill patients mechanically ventilated in the ICU experience frustration, anxiety, anger, fear, depersonalization, powerlessness, and a sense of futility when they have lost the ability to communicate via voicing. Additionally, these patients can experience depression and withdrawal from active participation in their medical care.¹⁸ Placing a tracheostomy can allow for less sedation and agitation, enhanced ability to produce speech, ability to take oral nutrition, increased patient comfort, ability to be transferred out of the ICU, and ease of replacement of tracheostomy tube.⁷

There are many other important clinical outcomes asides from mortality to consider when examining tracheostomy outcomes in MICU patients. These include time to ventilator liberation, time to decannulation, MICU length of stay, hospital length of stay, and discharge disposition. These are important both for managing patient family expectations as well as health care resources (Table 2). On average the patients in this current study remained ventilator dependent for two weeks after the tracheotomy and remained tracheostomy dependent for about one month after the tracheotomy. The average MICU stay was approximately one month, and the average hospital stay was approximately 40 days (Table 2). It can also be discussed with family that less than 25% of patients will be decannulated prior to discharge, and more than 50% of the patients will be discharged to a facility. It is important to note that BMI was not significantly associated with these data points.

As outcomes are increasingly tracked and used as a marker for both reimbursement and quality, mortality rates after tracheostomy in specific patient populations need to be reported in the literature. Additionally, reporting length of hospital stays, duration of tracheostomy dependence, and discharge disposition are important variables when analyzing utilization of health care resources. If there were strong clinical predictors of inpatient or one-year mortality, these clinical factors could be discussed with both the patients and their surrogate decision makers. However, both this current study and previous studies have found that there are few clinical features predictive of inpatient and one-year mortality.^19,20 As noted, the placement of tracheostomy is a significant portion of these patients’ care that helps improve quality of life during critical illness and recovery. The lack of association between obesity and mortality both in this study and other studies is useful information when counseling patients and their surrogate decision makers.¹⁹

The management of MICU patients requires complex multidisciplinary care. At our institution, the Otolaryngology-Head and Neck Surgery and the Pulmonary Critical Care services began collaborating in order to offer percutaneous tracheostomy to appropriately selected patients. No difference in mortality was found between the open versus percutaneous techniques (Table 3). Additionally, no major short- or long-term complications were encountered in the patients who underwent percutaneous tracheotomy. MICU patients requiring prolonged mechanical ventilation have an average MICU stay of about one month and have a high inpatient and one-year mortality rate. Understanding and communicating these outcomes with patient’s families can help the families make informed decisions during a difficult time.

Conclusions

The one-year mortality rate for MICU patients who underwent tracheotomy was 37% in this study. Age is significantly associated with one-year mortality, while sex, surgical technique, BMI, and APACHE II score are not significantly associated with one-year mortality. Communicating these outcomes with patient’s families during the informed consent discussion can help the families make the best decision on behalf of the patient.