Original Research—Laryngology and Neurolaryngology
Otolaryngology–
Risk Factors for Posttracheostomy
Tracheal Stenosis
Michael Li, MD1, Yin Yiu, MD1, Tyler Merrill, MD2,
Vedat Yildiz, MS3, Brad deSilva, MD1, and Laura Matrka, MD1
No sponsorships or competing interests have been disclosed for this article.
Abstract
Objective. To determine the incidence of posttracheostomy tra-
cheal stenosis and to investigate variables related to the patient,
hospitalization, or operation that may affect stenosis rates.
Study Design. A combined retrospective cohort and case-
control study.
Setting. Tertiary care academic medical center.
Subjects and Methods. A total of 1656 patients who under-
went tracheostomy at a tertiary care medical center from
January 2011 to November 2016 were reviewed for evi-
dence of subsequent tracheal stenosis on airway endoscopy
or computed tomography. Forty-three confirmed cases of
posttracheostomy tracheal stenosis (PTTS) were compared
with a subgroup of 319 controls. Factors including medical
comorbidity, type and setting of tracheostomy, and hospitali-
zation details were analyzed.
Results. Five-year incidence of PTTS was 2.6%. Obesity was
the sole demographic factor associated with stenosis.
Hospitalization-related variables associated with stenosis
included tracheostomy after 10 days of orotracheal intuba-
tion and endotracheal tube cuff pressure 30 mm H2O.
The surgical variables associated with higher rates of steno-
sis included percutaneous technique and insertion of an ini-
tial tracheostomy tube size .6. Bjork flap creation was
negatively associated with stenosis. In multivariable analysis,
obesity and insertion of tracheostomy tube size .6 were
identified as risk factors.
Conclusion. Greater than 10 days of orotracheal intubation
prior to tracheostomy and endotracheal tube cuff pressure
30 mm H2O were associated with greater rates of subse-
quent tracheal stenosis. The only patient-related factor
associated with tracheal stenosis was obesity. Surgical vari-
ables associated with increased rates of subsequent stenosis
included placement of a tracheostomy tube size .6, use of
percutaneous technique, and failure to create a Bjork flap.
Keywords
tracheal stenosis, tracheostomy, prolonged intubation, pro-
longed ventilation, open tracheostomy, percutaneous
Head and Neck Surgery
2018, Vol. 159(4) 698–704
Ó American Academy of
Otolaryngology–Head and Neck
Surgery Foundation 2018
Reprints and permission:
sagepub.com/journalsPermissions.nav
DOI: 10.1177/0194599818794456
http://otojournal.org
tracheostomy, subglottic stenosis, risks of tracheostomy,
predictors of prolonged intubation
Received October 29, 2017; revised June 13, 2018; accepted July 24,
2018.
T
racheostomy is a standardized and routinely per-
formed surgical procedure that is indicated for
airway obstruction or prolonged respiratory failure.1
As compared with orotracheal intubation, tracheostomy
improves patient communication and mobility and allows
for oral intake.2 Additionally, a meta-analysis comparing
early versus late tracheostomy, with 10 days as a cutoff,
found that patients receiving early tracheostomy had
decreased mortality rates and increased likelihood of being
transferred out of the intensive care unit.3 Studies also
showed that early tracheostomy decreases the incidence of
ventilator-associated pneumonia and reduces duration of
intensive care unit and mechanical ventilation.4,5 However,
tracheostomy is not without risk, with the most common
early complications being hemorrhage, mucus plugging, and
accidental dislodgment. Late complications include tra-
cheoinnominate fistula, mucus plugging, and tracheal steno-
sis, with current literature estimating the incidence of
posttracheostomy stenosis (PTTS) at 1.5% to 1.7%.6-8
Halum et al found obesity, defined as body mass index
.30, to be associated with increased rate of PTTS.7 In this
study, we aim to further define PTTS and to determine
which patient characteristics, hospitalization-related factors,
and surgical variables have an effect on airway stenosis
rates.
1
Department of Otolaryngology, The Ohio State University, Columbus,
Ohio, USA
2
College of Medicine, The Ohio State University, Columbus, Ohio, USA
3
Center for Biostatistics, The Ohio State University, Columbus, Ohio, USA
This article was presented at the 2017 AAO-HNSF Annual Meeting & OTO
Experience; September 10-13, 2017; Chicago, Illinois.
Corresponding Author:
Laura Matrka, MD, Ohio State University Wexner Medical Center Eye and
Ear Institute, 915 Olentangy River Rd Suite 4000, Columbus, OH 43212,
USA.
Email: laura.matrka@osumc.edu
Li et al 699

Table 1. Diagnoses of Patients Who Received ICD Diagnoses

Related to Tracheal Stenosis and Were Excluded.
Paent Variables
Exclusion Reason n Age
Sex
Preexisting tracheal stenosis without history of tracheostomy 41
Bleeding from tracheal site or mechanical complication 34 Race
Mucus plugging or blood clot 30 Body Mass Index
Total laryngectomy 23
External airway compression from cancer 22 Charlson Comorbidity Index
Internal airway obstruction from cancer 15 Alcoholism Diagnosis
Trach performed at outside hospital 14
Anatomic exclusions Smoking History
Stomal stenosis 18 Surgical Variables
Supraglottic stenosis 8
Posterior glottic stenosis 5 Surgeon specialty
Other Procedural locaon
Received negative workup 13
Vocal fold paralysis 10 Technique (open vs. percutaneous)
Tracheitis 7 Incision type (horizontal vs. veral)
Tracheal edema 4
Bronchomalacia 3 Tracheal spreader use
Trauma 2 Carlage resecon
Calcification 1
Fistula 1 Bjork flap
Total 251 Outer flange securing sutures use
Abbreviation: ICD, International Classification of Diseases. Tracheostomy tube size
Hospitalizaon Variables
Methods Prior orotracheal intubaon
Length of orotracheal intubaon
Approval was obtained from The Ohio State University
Human Subjects Institutional Review Board, and patients ETT size
who underwent tracheostomy from January 2011 through ETT cuff pressure
November 2016 were identified with the Current
Procedural Terminology code 31600. These charts were History of cricothyrotomy
searched electronically for codes based on the International Tracheostomy tube
Classification of Diseases (ICD; 9th or 10th revision) corre-
sponding to ‘‘tracheal stenosis,’’‘‘subglottic stenosis,’’‘‘tra- Tracheostomy tube cuff pressure
cheostomy complications,’’ or ‘‘other diseases of trachea Figure 1. List of variables collected. ETT, endotracheal tube.
and bronchus,’’ in addition to the umbrella ICD codes for
these diagnoses. Patients yielded from this search were
included in the PTTS group if a detailed chart review found
(1) a physician note documenting the diagnosis of stenosis Data were analyzed with SAS 9.3 (SAS Institute Inc,
and (2) imaging or endoscopic confirmation with broncho- Cary, North Carolina). Prior to analysis, data were examined
scopy or tracheoscopy during the 5-year study period. for outliers, and logarithmic transformation was performed
Patients receiving an ICD code but not confirmed to have for selected parameters due to the violation of normality
PTTS were excluded per the criteria listed in Table 1. assumption for data modeling. Descriptive statistics are
To create a control group, an additional 320 patients who reported as mean 6 SD or total number and percentage.
underwent tracheostomy but never received one of the The association between PTTS and the other patient charac-
aforementioned ICD diagnoses were randomly selected and teristics were explored with the x2 test for categorical vari-
reviewed for evidence of stenosis. Demographic variables ables or 2-sample t test for continuous variables. Data were
and information related to surgical service were collected further interrogated with univariate and multivariable logis-
for all 1656 patients who underwent tracheostomy. From tic regression analysis. Inclusion criteria was P \ .15 and
the medical records of confirmed patients with PTTS and exclusion criteria P . .20 for the multivariable model.
the 320 controls, we performed a detailed manual chart Significance level was set at a  0.05. An in-depth analysis
review (Figure 1). of the obese population was performed after obesity was
700 Otolaryngology–Head and Neck Surgery 159(4)

Table 2. Demographics. days previously and was found to have stenosis during revision
Tracheal Stenosis, n (%) or Mean 6 SD tracheostomy.
Of the 320 charts randomly chosen for review to serve as
Characteristics No Yes P Valuea the control group, 1 was identified as a duplicate with a sep-
arate medical record number and therefore excluded, and
n 319 43 none included PTTS, yielding 319 controls. The following
Age 57.7 6 14.1 60.9 6 13.6 .16 results correspond to these 319 controls and 43 patients with
Sex .40 PTTS.
Female 118 (90.1) 19 (10.7) Age, sex, race, Charlson Comorbidity Index, alcoholism,
Male 201 (90.9) 24 (13.9) and tobacco use were not associated with PTTS (Table 2).
Race .06 Mean body mass index was greater among patients with
Nonwhite 54 (85.7) 13 (19.4) PTTS (35.9 vs 29.8, P = .0014), who were also more likely
White 265 (91.7) 30 (10.2) to be obese (odds ratio [OR], 4.17; 95% CI, 2.06-8.43; P \
BMI 29.8 6 11.3 35.9 6 11.0 .0014 .0001). Obesity was strongly associated with PTTS in both
BMI category \.0001 univariate and multivariable analyses (P \ .0001); thus, the
Nonobese 197 (96.1) 12 (5.7) data were further analyzed to examine the relationship
Obese 122 (82.9) 31 (20.3) between them.
CCI 4.01 6 2.32 4.09 6 2.17 .82 Obese patients were more likely to be female (59.9% vs
Alcoholism .17 31.6%, P \ .0001), were less likely to have a history of
Yes 49 (94.2) 3 (5.8) alcoholism (6.8% vs 20.9%, P \ .0001), and had lower
No 259 (90.0) 39 (13.1) cumulative smoking pack-years (19.4 vs 28.3, P = .01).
Smoking .428 However, none of these variables were independent risk fac-
Current smoker 74 (89.1) 9 (10.9) tors for tracheal stenosis. There were no other significant
Former smoker 74 (86.1) 12 (13.9) differences in comorbidities or demographics between obese
Never smoker 63 (92.6) 5 (7.4) and nonobese patients.
Pack-years 25.2 6 29.8 21.4 6 24.4 .41 The only surgical variable that differed between obese
Abbreviations: BMI, body mass index; CCI, Charlson Comorbidity Index. and nonobese patients was size of tracheostomy tube placed
a
Bold indicates P \.05. at initial tracheostomy, which was larger for obese patients
(6.8 vs 6.4, P = .013). Obese patients were 25.4 times more
likely to have a tracheostomy tube with an extended proxi-
identified as a patient factor strongly associated with tra- mal portion inserted (95% CI, 7.69-84.0, P \ .0001). There
cheal stenosis. was no difference in rate of PTTS between obese patients
receiving and not receiving a tracheostomy tube with a
Results proximal extended portion (27.6% vs 27.7% P = .98). No
other surgical variables differed between obese and nonob-
Patient Characteristics ese patients.
A total of 1656 patients undergoing tracheostomy were identi- Maximum endotracheal tube (ETT) and tracheostomy
fied. Of these, 294 (17.9%) received an ICD code correspond- tube cuff pressures did not differ between obese and nonob-
ing to ‘‘subglottic stenosis,’’‘‘tracheal stenosis,’’‘‘tracheostomy ese patients, and the 2 groups did not differ in the propor-
complications,’’ or ‘‘other diseases of trachea and bronchus’’ tion with higher-than-acceptable cuff pressures (.30 mm
and were further reviewed for evidence of stenosis as H2O). The duration of preceding orotracheal intubation was
described in the Methods section. A total of 251 patients were greater among obese patients, although not significantly so
excluded (Table 1). Forty-three patients were confirmed to (13.7 vs 11.2 days, P = .089). However, obese patients were
have PTTS, yielding a 5-year incidence rate of 2.6%. Mean 2.23 times more likely to receive a late tracheostomy,
time from tracheostomy to diagnosis was 409 6 400 days. defined as occurring after 10 days of orotracheal intubation
Mean time from tracheostomy to death was 391 6 385 days (95% CI, 1.09-4.36; P = .023). Late tracheostomy itself was
(n = 121, 34%). Of the 43 patients confirmed to have PTTS, independently associated with tracheal stenosis (OR, 3.27;
16 were diagnosed via flexible tracheoscopy performed 95% CI, 1.14-9.40; P = .024).
through a tracheostoma, 13 via bronchoscopy, 9 via computed
tomography, 4 via direct laryngoscopy (1 with a video laryn-
goscope), and 1 during intraoperative revision tracheostomy. Surgical Variables
Of the 9 patients diagnosed via computed tomography, 4 had Between open and percutaneous tracheostomies, there was
their diagnoses later confirmed with flexible or direct laryngo- an increased risk of PTTS in the percutaneous tracheostomy
scopy. The patient initially diagnosed via video laryngoscope group (OR, 3.41; 95% CI, 1.40-8.33; P = .005); this type of
during intubation had the diagnosis subsequently confirmed tracheostomy was performed exclusively by nonotolaryngol-
via computed tomography 3 days later. The patient diagnosed ogists in our cohort. Patients receiving a percutaneous tra-
intraoperatively had undergone percutaneous tracheostomy 7 cheostomy were 3.4 times more likely to develop PTTS
Li et al
Table 3. Number of PTTS Cases by Specialty.a
ACS Burn
No PTTS 97 11
PTTS 2 0
Total 99 11
Abbreviations: ACS, acute care surgery; CTS, cardiothoracic surgery; GS, general surgery; OTO, otolaryngology; PTTS, posttracheostomy tracheal stenosis.
a
There was no difference in rate of stenosis based on stratified specialty. Overall, there was a difference in PTTS incidence between otolaryngologists and
nonotolaryngologists (2.1% vs 4.2%, P = .0312).
than those receiving an ‘‘ideal’’ tracheostomy: performed
open, with a Bjork flap, and not as a conversion from a cri-
cothyrotomy or percutaneous tracheostomy (95% CI, 1.13-10.0;
P = .0026). There were no demographic or hospitalization-
related differences between the ‘‘ideal’’ tracheostomy and percu-
taneous tracheostomy groups. However, between these sub-
groups, percutaneous tracheostomies were more likely to be
completed with a tracheostomy tube size .6 (OR, 6.07; 95%
CI, 2.52-14.6; P \ .001).
Creation of a Bjork flap was negatively associated with
PTTS (OR, 0.42; 95% CI, 0.20-0.88; P = .0024).
Otolaryngologists created a Bjork flap in 96% of cases, in
contrast to nonotolaryngologists, who utilized the technique
in 15% of cases (P \ .0001). In 11 cases, open surgical tra-
cheostomy was performed after an emergent cricothyrotomy
or failed percutaneous tracheostomy during the same hospi-
talization, and these patients were 6.14 more likely to
develop PTTS (95% CI, 1.69-22.24; P = .031).
Ninety-seven percent of surgeons utilized outer-flange
securing sutures, with no difference in stenosis rates from
those who did not. Otolaryngologists more frequently uti-
lized a horizontal incision on the trachea (P \ .0001) and
were less likely to resect cartilage (P \ .0001) or employ
use of a tracheal spreader (P \ .0001). Other than creation
of a Bjork flap, there was no difference in PTTS rate based
on these technical differences. Likewise, in examining all
1656 tracheostomies, PTTS incidence rate did not differ sig-
nificantly by stratified surgeon specialty (Table 3).
However, when groups were collapsed to compare otolaryn-
gologists and nonotolaryngologists, there was a lower inci-
dence of PTTS in the otolaryngologist group (2.0% vs
4.2%, P = .0312).
Hospitalization-Related Factors
Patients with PTTS were 2.8 times more likely to have been
orotracheally intubated during the same hospital stay prior
to tracheostomy (95% CI, 1.48-5.41; P = .002). There was a
trend toward a longer duration of preceding orotracheal
intubation for patients with PTTS (15.0 vs 12.1 days, P =
.07), who also were 3.3 times more likely to receive a late
tracheostomy, defined as .10 days of preceding orotracheal
intubation (95% CI, 1.14-9.40; P = .02). Orotracheal intuba-
tion with ETT size .7.5 was not identified as an indepen-
dent risk factor for PTTS (P = .43), and mean ETT size did
701
CTS OTO GS Other Total
147 1218 132 8 1613
6 26 8 1 43
151 1230 136 9 1656
not differ between patients with and without PTTS (7.64 vs
7.71, P = .62).
Orotracheally intubated patients who had the ETT cuff
inflated to .30 mm H2O were 2.89 times more likely to
develop PTTS (95% CI, 1.00-8.36; P = .04). However, there
was no association with tracheostomy tube cuff pressure
.30 mm H2O and PTTS (OR, 1.15; 95% CI, 0.38-3.47; P =
.80). Table 4 presents a summary of all factors associated
with PTTS.
Discussion
To our knowledge, we present the largest retrospective analy-
sis of PTTS in the United States. Tracheal stenosis is a chal-
lenging entity to treat and may require surgical intervention
ranging from balloon dilation to tracheal resection and reanas-
tomosis.9,10 Previous studies found tracheal stenosis to be the
most common long-term complication of tracheostomy, with
incidence rates of 1.7% and 1.8%.6,7 The 5-year incidence rate
in our cohort was comparable at 2.6%.
Obesity has been identified as a risk factor for postintu-
bation and posttracheostomy tracheal stenosis.7,11 In this
study, obese patients had longer durations of preceding oro-
tracheal intubation and were more likely to receive late tra-
cheostomy. These 2 factors were independently associated
with PTTS. Thus, it appears that a predisposition to pro-
longed orotracheal intubation prior to tracheostomy may
contribute to increased rates of PTTS in the obese popula-
tion. In addition, obese patients had larger tracheostomy
tubes inserted at the time of surgery, and tracheostomy tube
size .6 was associated with PTTS in the multivariable anal-
ysis. Larger tracheostomy tubes are unlikely to reduce aero-
dynamic resistance to a significant degree, as studies
demonstrated minimal difference in ventilator pressures
when comparing different ETT sizes.11-13 Furthermore, ana-
tomic studies found that tracheal diameter varies with
patient height rather than weight.11 Certainly, using a tra-
cheostomy tube size .6 may be unavoidable. But there is
no evidence to suggest a benefit for obese patients, and evi-
dence presented in this study suggests that placing an over-
sized tube may increase rates of PTTS. For obese patients,
it would instead be beneficial to place a tracheostomy tube
with an extended proximal limb; specifically, bypassing soft
tissue anterior to the trachea allows for better position of the
tip of the tube within the tracheal lumen.
702 Otolaryngology–Head and Neck Surgery 159(4)

Table 4. Variables Associated with Posttracheostomy Tracheal Stenosis in the 352-Patient Cohort.
Tracheal Stenosis, n (%) or Mean 6 SD Univariable Model Multivariable Model

Characteristics No Yes P Valuea OR (95% CI) P Valuea OR (95% CI)

BMI 29.8 6 11.3 34.8 6 10.4 .014 1.03 (1.00-1.06)

Obesity \.0001 6.14 (1.69-22.24) .0001 5.15 (2.23-11.87)
No 197 (96.1) 8 (3.9)
Yes 122 (82.9) 25 (17.1)
Early vs late tracheostomy .034 4.01 (1.09-14.7)
Early 49 (94.2) 3 (5.8)
Late 57 (80.3) 14 (19.7)
Previous orotracheal intubation .032 2.7 (1.05-4.45)
No 214 (93.1) 16 (6.9)
Yes 105 (88.1) 17 (13.9)
Initial tracheostomy size .011 2.5 (1.20-5.2) .013 2.60 (1.21-5.54)
6 237 (92.9) 18 (7.1)
.6 79 (84.0) 15 (16.0)
Max ETT cuff pressure .040 3.38 (1.00-11.3)
30 55 (90.1) 6 (9.9)
.30 19 (73.1) 7 (26.9)
Bjork flap .013 0.4 (0.19- 0.84)
No 73 (83.9) 14 (16.1)
Yes 246 (92.8) 19 (7.2)
Conversion tracheostomy .001 6.14 (1.69-22.24)
No 312 (91.5) 29 (8.5)
Yes 7 (63.6) 4 (36.4)
Ideal tracheostomy .022 3.36 (1.13 -10.0)
No 20 (80) 5 (25)
Yes 242 (93.1) 18 (6.9)

Abbreviations: BMI, body mass index; ETT, endotracheal tube; OR, odds ratio.
a
Bold indicates P \.05.

Ideally, ETT and tracheostomy tube cuffs should be
inflated to 20 to 30 mm H2O, as pressures .30 mm H2O
cause venous stasis, resulting in tissue necrosis and an
inflammatory response that may contribute to tracheal ste-
nosis.14 Our data further suggest that higher-than-acceptable
ETT cuff pressures increase not only a patient’s risk for
developing postintubation tracheal stenosis but also one’s
risk for PTTS.
Stenosis is also recognized as the most common long-term
complication following orotracheal intubation. Postintubation
airway stenosis incidence ranges from 10% to 22%, with 1%
to 2% of patients being symptomatic.15-17 This symptomatic
rate is comparable to the incidence of PTTS.7 Given that stud-
ies of PTTS have been retrospective chart reviews, which may
discover only symptomatic tracheal stenosis, it is possible that
rates of stenosis after tracheostomy and orotracheal intubation
are similar.
While PTTS and postintubation tracheal stenosis inci-
dence rates may not differ significantly, there are other ben-
efits of early tracheostomy versus prolonged orotracheal
intubation. Studies in the pulmonary and critical care litera-
ture showed that early tracheostomy facilitates ventilator
weaning, decreases rates of aspiration pneumonia, and
improves patient comfort. Papuzinski et al identified factors
predicting prolonged intubation and found patient age,
chronic pulmonary disease, ratio of arterial:inspired O2
\200, and hypernatremia to be predictors of prolonged intu-
bation.18 Clark et al developed a scoring system that pre-
dicted the need for prolonged intubation with 100%
specificity when patients met 4 criteria.19 Unfortunately,
this study did not include obesity in the analysis. Our data
suggest that obesity is associated with prolonged mechanical
ventilation and respiratory failure as an indication for tra-
cheostomy, both associated in turn with increased stenosis
rates. Thus, to reduce rates of PTTS, clinicians should fur-
ther favor early tracheostomy for obese patients for whom
prolonged mechanical ventilation is anticipated.
When a tracheostomy was performed, creation of a Bjork
flap was negatively associated with PTTS. This may be
related to more stable positioning of the tracheostomy tube
within the trachea, reducing shear force and mucosal
trauma. Halum et al theorized that usage of outer-flange
security sutures reduces the shear forces generated as tra-
cheostomy tubes slide against tracheal mucosa, as occurs
Li et al
during coughing, sneezing, or positional changes.7 A Bjork
flap sutured to the skin, with the skin sutured to the outer
flange of the tracheostomy tube, provides even greater stabi-
lity. Thus, during violent actions, such as sneezing or
coughing, the tube and trachea theoretically move as a unit,
rather than sliding against each other. This may reduce
shear force and further minimize mucosal injury and
inflammation.
PTTS incidence varied with type of tracheostomy and
specialty of surgeons, with percutaneous technique and non-
otolaryngologist surgeon conferring increased risk. In our
cohort, otolaryngologists did not perform any percutaneous
tracheostomies and uniformly created a Bjork flap. As these
variables were both associated with PTTS incidence, they
may be driving the decreased incidence of PTTS in tra-
cheostomies performed by otolaryngologists. Additionally,
there was an even greater difference in stenosis rates
between an ‘‘ideal’’ tracheostomy (performed with open
technique and creation of a Bjork flap) and a percutaneous
tracheostomy. When percutaneous tracheostomy is per-
formed, it may be difficult to place the tracheal incision
accurately, especially in those with poor surface landmarks,
such as obese patients or those who have undergone previ-
ous surgery. PTTS following percutaneous tracheostomy is
more likely to be subglottic in nature and involve tracheal
fracture or tracheomalacia.20 Case reports of complete lumi-
nal obstruction and creation of tracheoesophageal fistulae
were also reported.21 While percutaneous tracheostomy is a
safe technique, given the known challenge in certain patient
populations and the association with PTTS, it is critical that
clinicians be judicious in patient selection when performing
percutaneous tracheostomy and that a tracheostomy tube
size 6 be used when possible. It should be noted that
while percutaneous tracheostomy was more likely to be per-
formed with an initial tracheostomy tube size .6, both vari-
ables were independently identified as risk factors for
PTTS.
Although we present the largest data set of PTTS to date,
these data represent a single institution. However, given that
it is a high-volume tertiary care center, the results generated
from this study should be generalizable across similar aca-
demic centers. Nonetheless, prospective and multi-institutional
data could further strengthen the study. The overlap between
mean time from tracheostomy to death and mean time from
tracheostomy to diagnosis suggests that some patients may
have died prior to developing stenosis. While this is a chal-
lenge in all retrospective studies, a benefit of our study design
is that all 1656 patients who underwent tracheostomy were
queried for stenosis by ICD code. Furthermore, to our knowl-
edge, the 5-year study period in this study is the longest pub-
lished to date. As the mean time from tracheostomy to
diagnosis was 409 days in our population, this study may rep-
resent a more accurate characterization of PTTS than studies
with 1-year study periods.
Due to the retrospective nature of our study, not all
patients received bronchoscopy. Nevertheless, all PTTS
cases had either imaging- or bronchoscopy-confirmed
703
stenosis, and all patients who did not have bronchoscopy
had their stenosis directly visualized intraoperatively or by
tracheoscopy. Due to our study design and selection of ste-
nosis cases by ICD diagnosis code, some patients without
clinically apparent stenosis may have been omitted. The
likelihood of this appears low given that all patients without
these ICD codes (which composed our control group) were
free of stenosis. Furthermore, we sought to capture all
patients with PTTS by manual chart review for any patient
with an umbrella ICD code for tracheostomy complications,
including tracheal stenosis. Studies aimed at the underlying
molecular pathogenesis of stenosis are needed to address
why some patients develop severe or treatment-refractory
stenosis while others remain asymptomatic. We emphasize
the need for adequately powered prospective studies
designed to predict factors related to prolonged intubation
and better define the natural history of PTTS. In addition,
studies characterizing outcomes such as decannulation rate
and treatment patterns in PTTS are needed.
Conclusion
Obesity, insertion of a tracheostomy tube size .6, tra-
cheostomy performed after 10 days of orotracheal intuba-
tion, ETT cuff pressure .30-cm H2O, and percutaneous
technique are risk factors for development of PTTS.
Creation of a Bjork flap is associated with a lower rate of
PTTS. Obese patients are more likely to be intubated .10
days prior to tracheostomy and to receive larger tracheost-
omy tubes. PTTS incidence following early tracheostomy
appears to be comparable to that of postintubation stenosis,
with early tracheostomy providing the additional benefit of
reducing intensive care unit length of stay and facilitating
earlier return to normal diet. Thus, the recommendation for
early tracheostomy should be given serious consideration,
with particular attention to obese patients. As tracheal dia-
meter varies with patient height rather than weight, concerns
about inadequate ventilation with smaller tracheostomy
tubes in obese patients do not appear to be well supported
by the literature and may contribute to increased incidence
of PTTS.
Author Contributions
Michael Li, conception of project, data acquisition and analysis,
manuscript preparation, final approval, Yin Yiu, data acquisition
and analysis, manuscript preparation, final approval, Tyler
Merrill, conception of project, data acquisition and analysis,
manuscript preparation, final approval, Vedat Yildiz, data acquisi-
tion and analysis manuscript preparation, final approval, Brad
deSilva, data acquisition and analysis, manuscript preparation,
final approval, Laura Matrka, conception of project, manuscript
preparation, final approval.
Disclosures
Competing interests: None.
Sponsorships: None.
Funding source: None.
704 Otolaryngology–Head and Neck Surgery 159(4)

References 12. Weissman C. Flow-volume relationships during spontaneous

breathing through endotracheal tubes. Crit Care Med. 1992;20:
1. Jackson C. Tracheostomy. Laryngoscope. 1909;19:285-290.
615-620.
2. Heffner JE. Medical indications for tracheotomy. Chest. 1989;
13. Stenqvist O, Sonander H, Nilsson K. Small endotracheal tubes:
96:186-190.
ventilator and intratracheal pressures during controlled ventila-
3. Andriolo BN, Andriolo RB, Saconato H, Atallah ÁN, Valente O.
tion. Br J Anaesth. 1979;51:375-381.
Early versus late tracheostomy for critically ill patients. Cochrane
14. Totonchi Z, Jalili F, Hashemian SM, Jabardarjani HR.
Database Syst Rev. 2015;(1):CD007271.
Tracheal stenosis and cuff pressure: comparison of minimal
4. Siempos II, Ntaidou TK, Filippidis FT, Choi AMK. Effect of
occlusive volume and palpation techniques. Tanaffos. 2015;14:
early versus late or no tracheostomy on mortality and pneumo-
252-256.
nia of critically ill patients receiving mechanical ventilation: a
15. Zias N, Chroneou A, Tabba M, et al. Post tracheostomy and
systematic review and meta-analysis. Lancet Respir Med.
post intubation tracheal stenosis: report of 31 cases and review
2015;3:150-158.
of the literature. BMC Pulm Med. 2008;8:1-9.
5. Freeman B, Morris P. Tracheostomy practice in adults with
16. Stauffer J, Olson D, Petty T. Complications and consequences
acute respiratory failure. Crit Care Med. 2012;40:2890.
of endotracheal intubation and tracheotomy: a prospective
6. Goldenberg D, Ari E, Golz A, Danino J, Netzer A, Joachims
study of 150 critically ill adult patients. Am J Medicine. 1981;
H. Tracheotomy complications: a retrospective study of 1130
70:65-76.
cases. Otolaryngol Head Neck Surg. 2000;123:495-500.
17. Grillo HC, Donahue DM. Post intubation tracheal stenosis.
7. Halum S, Ting J, Plowman E, et al. A multi-institutional analysis
Semin Thorac Cardiovasc Surg. 1996;8:370-380.
of tracheotomy complications. Laryngoscope. 2012;122:38-45.
18. Papuzinski C, Durante M, Tobar C, Martinez F, Labarca E.
8. Kettunen W, Helmer S, Haan J. Incidence of overall complica-
Predicting the need of tracheostomy amongst patients admitted
tions and symptomatic tracheal stenosis is equivalent following
to an intensive care unit: a multivariate model. Am J
open and percutaneous tracheostomy in the trauma patient. Am
Otolaryngol. 2013;34:517-522.
J Surg. 2014;208:770-774.
19. Clark PA, Inocencio RC, Lettieri CJ. I-TRACH: validating a
9. Lorenz RR. Adult laryngotracheal stenosis: etiology and surgi-
tool for predicting prolonged mechanical ventilation [published
cal management. Curr Opin Otolaryngol Head Neck Surg.
online November 28, 2016]. J Intensive Care Med.
2003;11:467-472.
20. Raghuraman G, Rajan S, Marzouk J, Mullhi D, Smith F. Is tra-
10. Herrington H, Weber S, Andersen P. Modern management of
cheal stenosis caused by percutaneous tracheostomy different
laryngotracheal stenosis. Laryngoscope. 2006;116:1553-1557.
from that by surgical tracheostomy? Chest. 2005;127:879-885.
11. Schiff B. The relationship between body mass, tracheal diameter,
21. Roxbury C, Qualliotine J, Molena D, Kim Y. Unusual airway
endotracheal tube size, and tracheal stenosis. Int Anesthesiol Clin.
complication after percutaneous tracheotomy: case report and
2017;55:42-51.
literature review. Laryngoscope. 2015;125:1883-1885.