|
Alice C. Tzeng, M.D., Department of Physical Medicine
and Rehabilitation, University of Medicine and Dentistry
of New Jersey (UMDNJ)-New Jersey Medical School, Newark,
N.J. John
R. Bach, M.D., F.C.C.P., F.A.A.P.M.R., Professor
of Neurosciences, Department of Neurosciences, and Professor
of Physical Medicine and Rehabilitation, Department
of Physical Medicine and Rehabilitation, UMDNJ-New Jersey
Medical School, Newark, N.J.
This work was supported by grant H133 P 70011, Advanced
Multidisciplinary Fellowship in Rehabilitation Outcomes
Research to the Department of Physical Medicine and
Rehabilitation, UMDNJ-New Jersey Medical School, from
the National Institute on Disability and Rehabilitation
Research. The work was performed at University Hospital,
UMDNJ-New Jersey Medical School, Newark, NJ.
Corresponding Author:
Alice C. Tzeng, M.D.
Department of Physical Medicine and Rehabilitation
University Hospital B-403
UMDNJ-New Jersey Medical School
150 Bergen Street
Newark, N.J. 07103
phone: 1-973-972 4393; fax: 1-973-972 5725;
email: atzeng@umdnj.edu
Abstract
Study Objective: To evaluate the effects of a respiratory
muscle aid protocol on hospitalization rates for respiratory
complications of neuromuscular disease.
Design:
A retrospective cohort study.
Methods:
A home protocol was developed in which oxyhemoglobin
desaturation (dSaO2) was prevented or reversed by the
use of noninvasive intermittent positive pressure
ventilation (IPPV) and manually and mechanically assisted
coughing as needed. The patients (pts) who had more
than one episode of respiratory failure before having
access to the protocol were considered to have had pre-protocol
periods (Group 1). Other patients were given access
to the protocol when their assisted peak cough flows
(PCFs) decreased below 270 L/m before any episodes of
respiratory distress (Group 2). Hospitalizations (hosp)
and days hospitalized (days) were compared longitudinally
for pre-protocol and protocol access periods (Group
1). In addition, avoided hospitalizations were identified
as "episodes" of need for continuous
ventilatory support and dSaO2s reversed by assisted
coughing that were managed at home. Data were segregated
by access to protocol and by extent of baseline
ventilator use.
Results:
Of the 47 Group 1 patients with pre-protocol periods
who have subsequently had "episodes", 10 had
episodes before requiring ongoing ventilator use.
They had 1.060.84 pre-protocol hosp/yr/pt and 20.7636.01
days/yr/pt over 3.423.36 yrs/pt vs. 0.030.11 hosp/yr/pt
and 0.060.20 days/yr/pt with protocol use over 1.940.74
yrs/pt. Of these 47 Group 1 patients, 33 eventually
required part-time ventilatory aid and using the protocol
as needed had 0.080.17 hosp/yr/pt and 1.433.71 days/yr/pt
over 3.913.50 yrs/pt, as opposed to 1.401.96 hosp/yr/pt
and 20.1441.15 days/yr/pt pre-protocol and pre-ventilator
use over 5.896.89 yrs/pt. Twelve patients in Group 1
eventually required continuous noninvasive ventilation
and using the protocol as needed had 0.070.14 hosp/yr/pt
and 0.390.73 days/yr/pt over 5.355.10 yrs/pt by comparison
with 0.970.74 hosp/yr/pt and 10.398.66 days/yr/pt over
2.181.91 yrs/pt pre-protocol and pre-ventilator use.
For the 94 patients overall when having access to the
protocol, 1.020.99 hosp/yr/pt were avoided by 16 patients
before requiring ongoing ventilator use over a 2.191.84
year period, 0.991.12 hosp/yr/pt were avoided by 49
part-time ventilator users over 3.883.45 years,
and 0.800.85 hosp/yr/pt were avoided by 18 full-time
ventilator users over 6.545.00 years. All pre-protocol,
protocol rate comparisons were statistically significant
at P <0.004.
Conclusion:
Patients have significantly fewer hosp/yr and days/yr
when using the protocol as needed than without the protocol.
The use of inspiratory and expiratory aids can significantly
decrease hospitalization rates for respiratory complications
of neuromuscular disease.
Key words:
Neuromuscular Disease; Exsufflation; Mechanical ventilation;
Respiratory therapy.
Abbreviations:
ALS=amyotrophic lateral sclerosis; DMD=Duchenne muscular
dystrophy; EtCO2=end-tidal carbon dioxide tension; IPPV=intermittent
positive pressure ventilation; SaO2=oxyhemoglobin saturation;
PCF=peak cough flow; VC=vital capacity
Introduction
The great majority of neuromuscular disease morbidity
and mortality is due to respiratory muscle weakness.1
About 90% of episodes of respiratory failure occur during
otherwise benign upper respiratory tract infections
("chest infections") rather than from insidiously
progressive CO2 narcosis or other respiratory pathology.2
During chest infections already impaired pulmonary function
is further compromised by airway mucus accumulation,
fatigue, and worsening dysfunction of already weak inspiratory
and expiratory muscles.3 Thus, for conventionally managed
patients, chest infections result in repeated pneumonias,
hospitalizations, endotracheal intubations, and ultimately,
in tracheostomy and death.
Up to 24 hour use of noninvasive IPPV has prolonged
the survival of over 700 patients with neuromuscular
ventilatory failure.4 For noninvasive IPPV to do this,
however, the ability to clear the airway of secretions
is critical, especially during chest infections and
following surgical anesthesia. Peak cough flows (PCFs)
of at least 160 L/m are the minimum required in adults
to clear airway debris without resort to tracheal intubation
(Figure 1).5,6 In our experience, assisted PCFs of greater
than 160 L/m can be attained for the great majority
of adult patients with neuromuscular disease, except
for those with advanced bulbar amyotrophic lateral
sclerosis (ALS).7 We have found that when routinely
measured assisted PCFs are below 270 L/m, they are very
likely to decrease below 160 L/m during chest infections
and the likelihood of pneumonia and respiratory failure
increases greatly. Considering the fact that both hypercapnia
and airway mucus accumulation can cause dSaO2, we hypothesized
that if we could identify patients at high risk of developing
acute respiratory failure on the basis of low PCFs,
and train and equip the patients to prevent a sustained
decrease in SaO2 below 95%, pulmonary morbidity and
mortality could be prevented.
Patients and Methods
All of the patients in this study were referred to
our Jerry Lewis Muscular Dystrophy Association clinic
from neurologists who made the primary diagnoses, or
they referred themselves to us once learning about us
from other patients. Their diagnoses, confirmed in our
clinic by standard criteria,8 mean ages when first experiencing
respiratory distress, and pulmonary function are listed
in Table 1. All of the patients resided at home and
had dedicated family members or personal care attendants.
The patients were routinely evaluated for vital capacity
(VC) and maximum insufflation capacity (MIC) by spirometry
(Wright spirometer, Mark 14, Ferraris Development and
Engineering Co, Ltd, London), end-tidal CO2 (EtCO2)
(Microspan 8090 capnograph, Biochem International,
Waukesha, WI), oxyhemoglobin saturation (SaO2) (Ohmeda
Model #3760, Louisville, CO), and unassisted and
assisted PCFs (Access Peak Flow Meter, HealthScan, Inc.,
Cedar Grove, N.J.) (Figure 1). Once assisted PCFs were
determined to be below 270 L/m, the patients were trained
in receiving deep volumes of air via nose or mouth piece
and in manually and mechanically assisted coughing and
they became candidates for this study. Of the patients
so evaluated, 94 met the PCF criteria and had one or
more episodes requiring ventilatory support. Although
exclusion criteria for this study were substance
abuse and chronic lung disease as defined by long-term
radiographic abnormalities in conjunction with SaO2
chronically below 95% or ratio of forced expiratory
volume in 1 sec to forced VC 2 standard deviations less
than normal, no patients were excluded on the basis
of these criteria.
Once VCs were below 2000 ml or about 50% of predicted
normal the patients were trained in "air stacking"
consecutively delivered volumes of air delivered from
a manual resuscitator or volume cycled ventilator. The
patients received the air volumes via simple mouth pieces,9
nasal interfaces, or for those with weak lips and buccal
muscles, via an oral-nasal interface or lipseal
(Malincrodt, Pleasanton, CA) (Figure 2). They stacked
the consecutively delivered air volumes, holding them
with a closed glottis, until the lungs and chest walls
were as deeply insufflated as possible. The air
stacking capability or MIC was quantified spirometrically.
Patients who could air stack in this manner could also
use noninvasive IPPV, that is, receive IPPV via mouth
pieces and nasal interfaces (inspiratory muscle aids)
as ventilatory support,10 whereas, those not capable
of holding deeper volumes than their VCs (MIC=VC) have
more difficulty using noninvasive ventilation because
of air leakage.
Patients with symptoms suggesting hypoventilation,
elevated EtCO2, or periods of daytime SaO2 below 95%
underwent nocturnal SaO2 monitoring. With symptoms or
nocturnal SaO2 means below 95%, the patients had trials
of nocturnal nasal IPPV using a portable volume ventilator
(PLV-100, Respironics Inc., Murrysville, PA). The
patients continued to use nocturnal nasal IPPV when
they felt less fatigue or had relief of other symptoms
of chronic hypoventilation and when nocturnal mean SaO2
was shown to increase. Assist control mode at a rate
of 12 and delivered volumes of 800 to 1500 ml were used
for virtually all adolescent and adult patients. Rates
were increased and volumes decreased for young children.
With time, more than nocturnal use often became
necessary. Nocturnal to 16 hrs per day was considered
part-time, and greater than 16 hrs per day was considered
full-time use. Most patients used noninvasive IPPV for
daytime (continuous) ventilatory support for the first
time during chest infections and weaned from continuous
assistance (except during chest infections), often for
at least several years.
The patients were also trained in manually and mechanically
assisted coughing. Since normal cough volumes are 2.30.5
L11 and assisted cough flows have been reported to be
significantly increased by air stacking once VCs have
decreased below 1.5 L, to maximize assisted cough flows
the patients were told to air stack to deep lung volumes
once their VCs were less than 1500 ml.12 Thus, an assisted
cough most often consisted of insufflating patients
to approach their MICs and then applying an abdominal
thrust or ptussive squeeze (Figure 1), which ever produced
the highest assisted PCFs, timed to glottic opening.12
The patients were told to practice this technique with
care providers and their success was monitored by routinely
measuring assisted PCFs via a peak flow meter.
Mechanical cough assistance was provided by mechanical
insufflation-exsufflation (MI-E) with the use of an
In-exsufflator (J. H. Emerson Company, Cambridge, MA).12,13
The In-exsufflator provides an independently adjusted
deep insufflation via an anesthesia mask or a tube
if the patient is intubated. The In-exsufflator
insufflation pressure is usually set at +35 to +60 cm
H2O. Insufflation is followed by an exsufflation
provoked by an immediate pressure drop to -35 to -60
cm H2O. Each insufflation and exsufflation is usually
about 2 to 3 seconds. MI-E was particularly important
when assisted PCFs were marginal (about 160 to 270 L/min)
as bulbar muscle weakness prevented the retention of
deep volumes of air, and when scoliosis prevented optimal
manual thrusts. During the exsufflations abdominal
thrusts or ptussive squeezes were applied to further
increase exsufflation (cough) flows (Figure 3).
The patients received pulse oximeters. The protocol
consisted of using noninvasive IPPV continuously
or as needed to maintain eucapnia and normal SaO2, and
manually and mechanically assisted coughing as needed
to promptly reverse any decreases in SaO2 below 95%,
particularly during chest infections. They were told
that the dSaO2s would be due to some combination of
hypercapnia and, most commonly, airway mucus accumulation
and if these were not managed immediately the dSaO2s
would persist and atelectasis and pneumonia would
result. If not already using noninvasive IPPV or
MI-E, they were provided with rapid access (less than
2 hours) to a portable volume ventilator and an In-Exsufflator
(J. H. Emerson Co., Cambridge, MA).14 No treated patients
were symptomatic for under or over ventilation. Colds
often necessitated almost continuous care giver attention.
This was especially true during severe episodes during
which patients often slept poorly and required manually
and mechanically assisted coughing essentially
around-the-clock. No patients who were regularly evaluated
failed to be properly trained and equipped or refused
the protocol.
Conventional management was defined as any ambulatory
management not including continuous noninvasive IPPV
and mechanically assisted coughing with oximetry
feedback. Differences between conventional and
protocol approaches are summarized in Table 2. Patients
were considered in Group 1 when, managed conventionally,
they had one or more episodes of respiratory failure
before introduction of the protocol. Those who had two
or more episodes of respiratory failure had "pre-protocol
periods". For these patients, the hospitalization
rates and days, number of intubations, and days
intubated were quantitated during the pre-protocol period
beginning with the first episode, and compared with
those subsequently while having access to the protocol
but free of ongoing ventilator use, when requiring ongoing
part-time, and full-time ventilatory assistance, respectively
(Table 3). The Group 1 patients who had access to the
protocol immediately following the first episode of
respiratory failure had no "pre-protocol period."
Group 2 patients were those who were placed on the
protocol when their assisted PCFs decreased below 270
L/m and eventually had hospitalizations or "avoided
hospitalizations" for respiratory distress. Avoided
hospitalizations were tabulated for all patients (Table
4). They were defined as acute episodes of respiratory
distress and loss of autonomous ability to breathe during
chest infections relieved by continuous ventilator use
along with the use of assisted coughing and MI-E to
reverse dSaO2-associated mucus accumulation. For patients
already requiring continuous ventilator use, only
the acute need for increased use of assisted coughing
and MI-E to reverse dSaO2-associated mucus accumulation
during febrile upper respiratory tract infections was
the criterion for avoided hospitalizations.
A paired t-test was used to compare hospitalization
rates and days, number of intubations and intubated
days for the pre-protocol and protocol access periods.
The Wilcoxon Signed-Rank test was used for nonparametric
distributions. A p<0.05 was considered to represent
statistical significance.
Results
The 94 study patients included 68 males and 26 females.
They had a mean age of 31.09 (range = 2.5 to 73.5) years
at the most recent evaluation. All had sufficient bulbar
muscle function to permit speech and assisted PCFs to
exceed 160 L/m. EtCO2 levels were normal for all 24-hour
noninvasive ventilatory support users. No non-ventilator
users or part-time ventilator users had EtCO2 levels
over 50 cm H2O.
Seventy-one Group 1 patients had initial episodes of
respiratory failure at 23.8517.68 years of age, 4.205.66
years before being referred to us and having access
to the protocol at age 27.2418.84 years. The 23 Group
2 patients were put on the protocol at age 24.2914.91.
This was 1.591.13 years before their first potential
episode of respiratory failure.
There were 6 patients in Group 1 who were managed conventionally
up to an initial hospitalization for respiratory failure
and were then given access to the protocol. Their data
are analyzed separately because they had no pre-protocol
period to compare with protocol periods. All six began
the protocol along with nocturnal (part-time) ventilatory
aid. Of the 6, 4 avoided 0.950.84 hosp/yr over 4.053.27
yrs/pt (p=0.11) and the other 2 have had no chest infections
or episodes of respiratory distress since beginning
the protocol. Eighteen of the Group 1 patients have
not had any chest infections or episodes of respiratory
distress since learning the protocol.
The rates of hospitalizations, hospitalization days,
number of intubations, and intubation days for the remaining
47 Group 1 patients are noted in Table 3. Avoided hospitalizations
of both groups combined are summarized in Table 4. The
protocol was used by 11 Group 1 patients for a mean
of 0.981.57 years before requiring ongoing ventilatory
assistance, 2.902.45 years for 51 Group 1 patients while
using ongoing part-time ventilatory assistance,
and by 21 Group 1 patients using ongoing full-time ventilatory
assistance for 4.354.81 years. The protocol was used
by 3 Group 2 patients 1.432.37 years before requiring
ongoing ventilatory assistance, 22 Group 2 patients
using part-time ventilatory assistance for 3.612.88
years and 9 Group 2 patients using full-time ventilatory
support for 5.294.88 years. The p-values as determined
by t-test compared the actual hospitalization rates
with those that would have occurred with conventional
management.
One 15 year old DMD patient with unassisted PCFs of
250 L/m and assisted PCFs of 300 L/m did not meet the
criteria for training and inclusion in this study.
One month later he developed a chest infection, pneumonia,
and was hospitalized. The protocol was instituted. He
did not require intubation and was discharged following
a hospitalization of 8 days.
Discussion
This study had no true control group. However, it would
be unethical to compare the hospitalization rates of
an untreated group with the hospitalizations avoided
by treatment. Conventional management includes
the hospitalization and invasive respiratory management
of any patient requiring ventilatory support with
no ventilator-free breathing ability, whereas, for us
the need for continuous ventilator use was a key criterion
for establishing an "avoided hospitalization."
At least while they were free of chest infections,
assisted PCFs greater than 160 L/min were attainable
for all the protocol-using patients throughout the study
period. This was true despite the fact that 13 of these
patients went on to require full-time noninvasive
ventilatory support and one patient had a VC below 100
ml. Thus, all of these patients were candidates
for long-term management of ventilatory muscle failure
without resort to tracheostomy.
The use of assisted PCF below 270 L/m as a threshold
criterion for introduction of this protocol appears
to be reasonable because we have only one patient who
developed pneumonia despite having assisted PCFs greater
than this. Further, it was about 1 1/2 years from
learning the protocol until needing to use it on the
average. Earlier introduction of the protocol would
further increase up front expenses. It is also possible
that other patients with PCFs greater than 270 L/m developed
respiratory failure and never returned to our clinic.
The fact that assisted PCFs can exceed 270 L/m when
patients are well does not mean that they will exceed
160 L/m or be adequate to clear airway secretions
during chest infections (thus the great utility of mechanical
insufflation-exsufflation for many), nor is it a guarantee
that the patients will have access to assisted coughing
when and as often as they need it during infections.
All of the patients in this study had very dedicated
and capable care providers who assisted coughing during
chest infections, at times every 5 to 10 minutes,
essentially around the clock. Patients without dedicated
and effective care providers may not succeed in avoiding
conventional invasive management.
Prior to initiating this protocol, we informed patients
to contact us at the first sign of a chest infection
and we would provide them with an oximeter and respiratory
muscle aids to avert hospitalization. However, two consecutive
patients presented to us only after they had already
developed severe dSaO2 and pneumonia and they required
hospitalization and were intubated. We, therefore, feel
that an oximeter in the home is especially important
for immediate feedback to the patient during chest infections.
Pneumonia and need for intubation are very unlikely
when the SaO2 baseline is maintained above 94% without
supplemental oxygen.
Although the delivered volumes we used may appear to
be high, the efficacy of noninvasive ventilation depends
on the alertness of the central nervous system during
sleep to decrease insufflation leakage to avoid asphyxia
and hypercapnia, and to obstruct the airway and increase
leakage to decrease insufflation volumes to avoid hyperventilation.4
Although we monitored EtCO2 during sleep in the past,15,16
we no longer feel that this is necessary. As noted in
the Results, all treated patients were asymptomatic
for under or over ventilation and all maintained essentially
normal EtCO2 levels. For patients using only nocturnal
noninvasive IPPV, daytime hypercapnia was usually mild
and well-tolerated when SaO2 remained within normal
limits.
Many protocol patients first used noninvasive IPPV
during chest infections and weaned to ongoing nocturnal
use of noninvasive IPPV after their first averted
hospitalizations. Oximetry was useful in guiding the
eventual ongoing daytime need and nocturnal use
of noninvasive IPPV.14,16,17 Non-protocol patients tended
to have repeated episodes of respiratory failure until
being introduced to the protocol or undergoing tracheostomy.
Although both tracheostomy and noninvasive respiratory
aids can prolong life, tracheostomy is associated with
numerous complications,18 an initial mean hospital stay
of 72 days, and significantly more hospitalization days
both acutely and long-term than is noninvasive management.2
It might be that the significantly lower hospitalization
rates of noninvasive vs. tracheostomy mechanical ventilation
users is due to reduced pulmonary inflammation,
maintenance of physiologic airway defense mechanisms,
and maintenance of pulmonary health long-term by avoiding
invasive respiratory interventions. Further, patients
almost invariably prefer noninvasive aids over tracheostomy
for safety, convenience, appearance, comfort, facilitating
effect on speech, sleep, swallowing, and general acceptability.19
Noninvasive aids should also, therefore, be preferred
by clinicians when effective.
Our patients were trained in the use of respiratory
muscle aids including volume-cycled ventilators in the
clinic and home settings. Unlike the pressure-cycled
machines conventionally used for nocturnal nasal IPPV,
the volume-cycled machines permitted the air stacking
necessary to maximize cough flows. Fourteen noninvasive
IPPV users went on to require definitive 24 hour ventilatory
support using volume-cycled ventilators without as yet
ever being hospitalized. Thus, provided that proper
attention is given to assisting cough, this study
demonstrates that even long-term and full-time need
for ventilatory support can be safely provided noninvasively.
References
1. Bach JR. Conventional approaches to managing neuromuscular
ventilatory failure. In: Bach JR, ed. Pulmonary Rehabilitation:
The Obstructive and Paralytic Conditions. Philadelphia:
Hanley & Belfus, 1996; 285-301
2. Bach JR, Rajaraman R, Ballanger F, et al. Neuromuscular
ventilatory insufficiency: the effect of home mechanical
ventilator use vs. oxygen therapy on pneumonia and hospitalization
rates. Am J Phys Med Rehabil 1998;77:8-19
3. Mier-Jedrzejowicz A, Brophy C, Green M. Respiratory
muscle weakness during upper respiratory tract infections.
Am Rev Respir Dis 1988; 138:5-7
4. Bach JR. Prevention of morbidity and mortality with
the use of physical medicine aids. In: Bach JR, ed.
Pulmonary rehabilitation: the obstructive and paralytic
conditions. Philadelphia: Hanley & Belfus, 1996;
303-29
5. Bach JR. Amyotrophic lateral sclerosis: predictors
for prolongation of life by noninvasive respiratory
aids. Arch Phys Med Rehabil 1995; 76:828-32
6. Bach JR, Saporito LR. Criteria for extubation and
tracheostomy tube removal for patients with ventilatory
failure: a different approach to weaning. Chest 1996;
110:1566-71
7. Kang SW, Bach JR. Maximum insufflation capacity:
the relationships with vital capacity and cough
flows for patients with neuromuscular disease. Am J
Phys Med Rehabil (in press)
8. Emery AEH. Diagnostic criteria for neuromuscular
disorders. Baarn, The Netherlands: European Neuromuscular
Centre, 1994
9. Bach JR, Alba AS, Saporito LR. Intermittent positive
pressure ventilation via the mouth as an alternative
to tracheostomy for 257 ventilator users. Chest 1993;
103:174-82
10 Bach JR. Update and perspectives on noninvasive
respiratory muscle aids: part 1--the inspiratory muscle
aids. Chest 1994; 105:1230-40
11. Leith DE. Cough. In: Brain JD, Proctor D, Reid
L, eds. Lung biology in health and disease: respiratory
defense mechanisms, part 2. New York: Marcel Dekker,
1977; 545-92
12. Bach JR. Mechanical insufflation-exsufflation:
comparison of peak expiratory flows with manually assisted
and unassisted coughing techniques. Chest 1993; 104:1553-62
13. Bach JR. Update and perspectives on noninvasive
respiratory muscle aids: part 2--the expiratory muscle
aids. Chest 1994; 105:1538-44
14. Hill NS, Eveloff SE, Carlisle CC, et al. Efficacy
of nocturnal nasal ventilation in patients with
restrictive thoracic disease. Am Rev Respir Dis
1992; 145:365-71
15. Bach JR, Alba AS. Sleep and nocturnal mouthpiece
IPPV efficiency in post-poliomyelitis ventilator
users. Chest 1994; 106:1705-10
16. Bach JR, Alba AS. Management of chronic alveolar
hypoventilation by nasal ventilation. Chest 1990;
97:52-57
17. Jimenez JFM, Sanchez de Cos Escuin J, Vicente CD,
et al. Nasal intermittent positive pressure ventilation:
analysis of its withdrawal. Chest 1995; 107:382-88
18. Bach JR, Intintola P, Alba AS, et al. The ventilator-assisted
individual: cost analysis of institutionalization versus
rehabilitation and in-home management. Chest 1992; 101:26-30
19. Bach JR. A comparison of long-term ventilatory
support alternatives from the perspective of the
patient and care giver. Chest 1993; 104:1702-06
|
Selected
Studies
Printer
Friendly Page
