LETTERS

Table 2. Results of multiple logistic regression analysis to predict in-hospital mortality among 821,902 patients based on sex, age,
race/ethnicity, year, socioeconomic status, history of diabetes mellitus, hospital type, and expected primary payer

Variables Parameter Standard Wald x2 P Value Odds 95% Confidence 95% Confidence
Estimates (b) Errors Ratio Interval (Lower) Interval (Upper)

Intercept 21.737 0.022

Male (reference) 1
Female 20.013 0.005 7.07 0.008 0.99 0.98 1.00
Age 0.020 0.000 11076 ,0.001 1.02 1.01 1.04
Race/ethnicity
White (reference) 505 1
Black 0.118 0.007 280 ,0.001 1.13 1.11 1.14
Hispanic 0.144 0.009 284 ,0.001 1.16 1.14 1.18
Asian or Pacific islander 0.126 0.015 75 ,0.001 1.13 1.10 1.17
Native American 0.073 0.032 5 0.03 1.08 1.01 1.15
Year
2009 (reference) 835 1
2010 20.056 0.007 56 ,0.001 0.95 0.93 0.96
2011 20.140 0.007 371 ,0.001 0.87 0.86 0.89
2012 20.189 0.007 692 ,0.001 0.83 0.82 0.84
Socioeconomic status
represented in quartiles, %
Quartile 1 (reference) 52 1
Quartile 2 20.036 0.007 26 ,0.001 0.97 0.95 0.98
Quartile 3 20.049 0.007 47 ,0.001 0.95 0.94 0.97
Quartile 4 20.024 0.007 10 ,0.001 0.98 0.96 0.99
Type of Hospital
Rural vs urban 0.040 0.012 2713 ,0.001 1.04 1.02 1.07
Nonteaching vs teaching 20.116 0.005 486 ,0.001 0.89 0.88 0.90
Primary expected payer
Medicare 707 1
Medicaid 0.107 0.009 149 ,0.001 1.11 1.09 1.13
Private Insurance 0.075 0.007 105 ,0.001 1.08 1.06 1.09
Self-pay 0.334 0.013 632 ,0.001 1.40 1.36 1.43
No charge 0.294 0.04 43 ,0.001 1.34 1.23 1.46
Other 0.035 0.017 4.2 0.04 1.04 1.002 1.07

Excessive Dynamic Airway Collapse: Fact, Fiction,
or Flow Limitation
To the Editor:
The recent paper by Weinstein and colleagues (1) with
commentary by Murgu and Stoy (2) reported on the observation
of dynamic airway collapse, referred to by the term excessive
dynamic airway collapse (EDAC), as evidence for tracheal
pathology.
Interest in tracheal collapse has increased in recent years
because of the advent of fast computed tomographic (CT) chest
imaging, which periodically reveals marked invagination of the
posterior tracheal membrane during expiration. However,
physiologists have described this phenomenon for many years (3).
The posterior membrane of the trachea invaginates during
breathing because of changes in transmural pressure (intraluminal-
surrounding pressure) (3, 4). During forced expiration, increases in
pleural pressure and decreases in intraluminal pressure combine to
decrease the transmural pressure applied to the central airways. The
resultant tracheal narrowing is directly related to the transmural
pressure (4).
In life, dynamic narrowing of the central airways occurs
during flow limitation (5, 6) and can be seen during tidal breathing
if the subject is breathing on his or her maximum flow-volume
curve. Tracheal narrowing is enhanced in patients with obvious small
airways disease such as chronic obstructive pulmonary disease, and
in subjects without lung disease but who are breathing at reduced
lung volumes with reduced pressure of elastic recoil (e.g., the obese).
Tracheas in these individuals are almost certainly structurally normal.
The reason for the narrowing is that circumstances in the
airway allow lower than normal transmural pressures to be
generated. For example, during expiration in chronic obstructive
pulmonary disease, expiratory flow is low, and the patient has
enough time to contract the muscles of expiration to raise the
pleural pressure higher than normal. At the same time, increases
in small airways resistance and/or low elastic recoil pressures
lower intraluminal pressures more than normal.
Weinstein and colleagues (1) detected tracheal collapse
during exercise in soldiers with unexplained dyspnea. Careful
workups did not find small airways disease, and the investigators
concluded that central airway collapsibility must be at fault.
Weinstein and collaborators (1) and Murgu and Stoy (2) state that
narrowing .50% is “excessive,” and they indict the trachea.
Alternatively, one might conclude that distal airways pathology
was present but simply could not be detected and that lower
transmural pressures during exercise are a more reasonable
explanation for the radiologic findings.

Letters 301

Weinstein and coworkers (1) and Murgu and Stoy (2)
acknowledge that flow limitation can cause collapse. However,
because Weinstein and colleagues failed to find evidence of
distal airways pathology, they invoked the new diagnosis, EDAC.
They reasoned that the narrowed trachea is the site of increased
resistance and the cause of dyspnea. However, the fact that they could
not detect subtle abnormalities in distal airways does not mean that
none were present. One might argue that the presence of flow
limitation (collapsing airways) in these patients implies the presence
of small airways disease rather than a primary tracheal abnormality.
Differences in interpretation of the nature of tracheal collapse
are of practical clinical importance. Some interventional
bronchoscopists call tracheal narrowing a disease, and they advocate
tracheal stenting to treat the obstruction. However, dynamic airway
collapse of the trachea, although dramatic and obvious on CT
imaging or bronchoscopy, does not cause airway obstruction (5, 6).
Studies of flow limitation define the factors limiting flow as the
resistance in small airways and airway properties just upstream of
the narrowed segment (6); that is, a choke point forms when the
intraluminal pressure drop equals the critical pressure of the airway
and that defines maximal flow. The airways downstream of the
choke point (e.g., between the choke point and the mouth, often
including main stem bronchi and trachea) can collapse (because
of transmural pressures) but the pressure drop (and resistance)
measured along the collapsed trachea are negligible (6, 7) and, in
classic teaching, this segment of the airway does not affect flow.
This reasoning should cause one to question the validity of the
diagnosis of EDAC as a pathological entity and also to question
the concept of tracheoplasty and stenting to treat this so-called disorder.
In the 1970s and 1980s, physiologists realized that stenting the
flow-limiting airways would be futile because the choke point will
simply move toward alveoli and form just upstream of any stent.
Reply
From the Authors:
We appreciate the comments of Dr. Smaldone regarding the recent
publication by Weinstein and colleagues, “Exercise-induced
excessive dynamic airway collapse in military personnel,” (1) and
the accompanying editorial by Murgu and Stoy (2). Dr. Smaldone
eloquently summarized the Starling resistor model and equal
pressure point theory explaining expiratory flow limitation.
We agree that excessive dynamic airway collapse (EDAC), as
seen on bronchoscopy or dynamic computed tomographic (CT)
imaging, occurs at the compressible segment of the central airway
downstream from the equal pressure point. We agree that tracheal
collapse is caused by increased airway resistance upstream from
the equal pressure point, decreased lung elastic recoil, and increased
pleural pressure. That mechanism explains the finding of EDAC
in patients with uncontrolled asthma, chronic obstructive
pulmonary disease (COPD), and obesity, especially when two or
more of those disorders coexist (3). Indeed, EDAC can be seen
in 20% patients with severe COPD when assessed by dynamic
chest computed tomography imaging (4).
We do not claim that EDAC is flow limiting. Even when EDAC
is defined conservatively as a forced expiratory collapse of .80%
302
LETTERS
Careful review of interventional trials (referenced by Weinstein
and coauthors [1]) testing various forms of tracheal stents and
stiffeners confirmed this prediction, because those maneuvers
failed to demonstrate any improvement in FEV1 or expiratory
flow. Tracheal pathology that can truly affect flow is a fixed
stenosis, serving as a fixed resistor during both phases of
respiration. Dynamic collapse during expiration as seen on CT or
bronchoscopy is an unlikely cause of obstruction.
Author disclosures are available with the text of this letter at
www.atsjournals.org.
Gerald C. Smaldone, M.D., Ph.D.
State University of New York at Stony Brook
Stony Brook, New York
References
1 Weinstein DJ, Hull JE, Ritchie BL, Hayes JA, Morris MJ. Exercise-
associated excessive dynamic airway collapse in military personnel.
Ann Am Thorac Soc 2016;13:1476–1482.
2 Murgu S, Stoy S. Excessive dynamic airway collapse: a standalone cause
of exertional dyspnea? Ann Am Thorac Soc 2016;13:1437–1439.
3 Macklem PT, Mead J. Factors determining maximum expiratory flow in
dogs. J Appl Physiol 1968;25:159–169.
4 Knudson RJ, Knudson DE. Pressure-flow relationships in the isolated
canine trachea. J Appl Physiol 1973;35:804–812.
5 Pride NB, Permutt S, Riley RL, Bromberger-Barnea B. Determinants of
maximal expiratory flow from the lungs. J Appl Physiol 1967;23:646–662.
6 Smaldone GC, Bergofsky EH. Delineation of flow-limiting segment and
predicted airway resistance by movable catheter. J Appl Physiol
1976;40:943–952.
7 Smaldone GC, Smith PL. Location of flow-limiting segments via airway
catheters near residual volume in humans. J Appl Physiol (1985)
1985;59:502–508.
Copyright © 2017 by the American Thoracic Society
of the airway lumen, there is no significant correlation between
end-expiratory or dynamic expiratory collapse and percentage
predicted FEV1 (5). Moreover, clinicians have documented
improvement in EDAC after lung transplantation or lung volume
reduction surgery in patients with emphysema, consistent with the
hypothesis that EDAC is a consequence of peripheral airway
disorders, and not central airway pathology.
The editorialists do not recommend stenting or membranous
tracheoplasty for patients with EDAC (5). We distinguish this
entity from tracheobronchomalacia, so that patients are not
inappropriately referred for stent insertion or tracheoplasty if their
sole symptom is dyspnea.
We believe that EDAC should also be distinguished
from the minimal (usually ,50%) dynamic airway compression
of the lower trachea that is seen in normal individuals.
Even though it may not be flow-limiting, EDAC has been
reported to be associated with worse quality of life (St. George’s
Respiratory Questionnaire score) in cigarette smokers, and
increased frequency and severity of COPD exacerbations (6).
For highly selected patients with EDAC, stenting or tracheoplasty
may improve quality of life (7, 8). This may be explained by
improvement in secretion clearance and cough or by improvement
in dyspnea by mechanisms not detected by pulmonary function
testing.
AnnalsATS Volume 14 Number 2 | February 2017