| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Guest Access | Sign In via User Name/Password
|
|
|
|||||||
Guest Access | Sign In via User Name/Password |
|||||||||
* From the Respiratory Intensive Care Unit (Drs. Corrado, Villella, and Gorini), Azienda Ospedaliera di Careggi Firenze, Florence, Italy; the Respiratory Intensive Care Unit (Drs. Confalonieri and Della Porta), Aziende Ospedaliere di Crema, Crema, Italy; and the Respiratory Intensive Care Unit (Dr. Mollica), Aziende Ospedaliere di Ospedale Forlanini Roma, Rome, Italy; and the Respiratory Intensive Care Unit (Dr. Marchese), Ospedale Civico, Palermo, Italy.
Correspondence to: A. Corrado, MD, FCCP, Unita’ di Terapia Intensiva Polmonare e, Fisiopatologia Toracica, Azienda Ospedaliera di Careggi, CTO, Largo Palagi 1, I-50136 Florence, Italy; e-mail: acorrado{at}qubisoft.it
 |
Abstract |
|---|
 TOP Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Design: A retrospective case-control study.
Setting: Four Italian respiratory intermediate ICUs.
Patients: Of a total of 393 COPD patients admitted to the ICU in 1996, 53 pairs were treated with the iron lung (NPV group). Patients treated with NIPPV (NIPPV group) were matched according to mean (± SD) age (70.3 ± 7.1 vs 70.3 ± 6.9 years, respectively), sex, causes of acute respiratory failure (ARF), APACHE (acute physiology and chronic health evaluation) II score (22.4 ± 5.3 vs 22.1 ± 4.6, respectively), pH (7.26 ± 0.05 vs 7.27 ± 0.04, respectively), and PaCO2 (88.1 ± 11.5 vs 85.1 ± 13.5 mm Hg, respectively) on admission to the ICU. The effectiveness of matching was 98.4%.
Results: Five patients from the NPV group (9.4%) and seven patients from the NIPPV group (13.2%) needed endotracheal intubation (EI). The treatment failure rate (ie, death and/or need of EI) was 20.7% in the NPV group and 24.5% in the NIPPV group (difference was not significant). The mean duration of mechanical ventilation (29.6 ± 28.6 vs 62.3 ± 35.7 h, respectively) and length of hospital stay (10.4 ± 4.3 vs 15 ± 5.2 d, respectively) among the 35 concordant surviving pairs were significantly lower in the NPV group than in the NIPPV group (p = 0.001 and p = 0.001, respectively).
Conclusions: These data suggest that both ventilatory techniques are equally effective in avoiding EI and death in COPD patients with ARF. Prospective trials are needed to confirm these preliminary results.
Key Words: acute respiratory failure • COPD • negative pressure ventilation • noninvasive positive-pressure ventilation • respiratory intermediate ICU
|
Introduction |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
It has been reported also, however, that COPD patients with severe respiratory acidosis and hypercapnic coma were successfully treated with a noninvasive ventilatory technique using negative-pressure ventilation (NPV).5 Furthermore, a case control study suggests that NPV is as effective as conventional mechanical ventilation in the treatment of ARF in patients with COPD and that it is associated with a shorter duration of ventilation and a similar length of hospital stay compared with conventional mechanical ventilation.6 This suggests that NPV may be superior to NIPPV in some cases7 and may be considered as a rescue technique when NIPPV fails.8 However, direct comparisons between the two noninvasive ventilatory techniques in the treatment of COPD patients with ARF are lacking.
In the present study, we compared COPD patients with ARF who were treated with NPV and NIPPV in four respiratory intermediate ICUs (RIICUs). The aim of this retrospective case-control study was to investigate whether there were differences in outcome and complications that could be related to the different ventilatory treatments employed.
|
Materials and Methods |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The level of assistance (ie, the nurse/patient ratio) was similar among the four settings (1:3 to 1:4).
Definitions
Cases (NPV Group):
Patients were admitted to the RIICU according to the following inclusion criteria: presence of COPD, presence of ARF, and the need for noninvasive mechanical ventilation.
The diagnosis of COPD was established on the basis of clinical history, physical examination, and the findings of the chest radiograph. Additional information was obtained from functional measurements when available 1 to 12 months before patients experienced ARF.
ARF was defined as a condition characterized by at least three of the following criteria: acute worsening of dyspnea; signs of right heart failure (ankle edema); severe hypoxemia (ie, PaO2/fraction of inspired oxygen [FIO2] ratio, < 240); and decompensated respiratory acidosis (ie, PaCO2, > 70 mm Hg; pH, < 7.30).
The ventilatory treatment was applied if, after 2 to 3 h of intensive pharmacologic treatment and optimum oxygen administration, there was no improvement in arterial blood gas levels and pH.
The exclusion criteria were the following: (1) postoperative conditions; (2) restrictive disorders (eg, kyphoscoliosis, neuromuscular disorders, or fibrothorax) as the cause of chronic respiratory failure; (3) neurologic disorders unrelated to hypercapnia or hypoxemia; (4) acute myocardial infarction; (5) left heart failure; (6) ARDS; (7) acute and chronic renal failure; (8) neoplasia; and (9) pulmonary embolism.
Control Subjects (NIPPV Group):
A control subject was
defined as a patient who had characteristics that were similar to those
of a patient in the NPV group (ie, a case). The
variables used for matching individual patients of NPV group were the
following: age (± 5 years); sex; causes triggering ARF; acute
physiology and chronic health evaluation (APACHE) II
score9
calculated within the first 24 h after
admission to the hospital (± 5 points); pH on hospital admission
(± 0.03); and PaCO2 on hospital admission
(± 6 mm Hg of the value for the case when that value was < 70 mm
Hg, and within 12 mm Hg when the value was 
70 mm Hg). When more
than one potential control subject was well-matched to a case, the
control subject with data closest to the case was selected.
The comparability of the two groups was further evaluated on the basis of the following data at hospital admission: PaO2/FIO2 ratio, Glasgow coma score,10 and serum level of albumin. Moreover, the plasma bicarbonate levels and pH value at the time of hospital discharge, the percentage of cor pulmonale, comorbidities, previous episodes of ARF, and the need for long-term oxygen therapy (LTOT) and home mechanical ventilation were analyzed as further criteria of comparability. The last five variables referred to the chronic health status of the patients before RIICU admission.
Modalities of Treatment and Outcome Assessment
All patients received oxygen therapy in order to obtain a
PaO2 of between 60 and 70 mm Hg, as
well as therapy with standard drugs (ie, inhaled
ß2-agonists, cardiokinetic agents, IV
aminophylline, IV diuretics, IV steroids, and IV antibiotics). When
the patients admitted to Careggi and Forlanini Hospitals (the NPV
group) did not respond to medical treatment, they were given a
first-line treatment with NPV provided by an iron lung (models CZ800
and C 900; Coppa Co; Biella, Italy), whereas the patients admitted to
Crema and Palermo Hospitals (the NIPPV group) were treated with NIPPV
(BiPAP; Respironics; Monroeville, PA; and models 7200 and 335
ventilators; Puritan Bennett; Pleasanton, CA). In these
patients, the following ventilatory modalities were applied:
pressure-assisted control ventilation (36 patients) and pressure
support ventilation (17 patients). Face and nasal masks were used in 36
and 17 patients, respectively. The mean (± SD) pressure delivered
during inspiration and expiration was 20 ± 4 and 5 ± 2 cm
H2O, respectively.
The iron lung settings for the NPV group were similar to those previously reported.5 The ventilator was set to deliver pressures ranging from -30 to -40 cm H2O (negative pressure) and from 10 to 15 cm H2O (positive pressure), respectively. The frequency was set to 15 cycles per minute, and the ratio of inspiratory time to total breathing cycle time (TI/TTOT) was set to 30% in patients with bradypnea (spontaneous frequency, < 10 cycles per minute). In the other patients, the frequency and the TI/TTOT ratio were individually set according to the spontaneous respiratory frequency in order to facilitate the adaptation of the patient to the ventilator. Oxygen was provided by nasal cannula or Venturi mask to increase the PaO2 level to between 60 and 70 mm Hg. To prevent upper airway obstruction due to the collapse of the tongue, an oropharyngeal airway was used in patients who presented with a deterioration of consciousness.
In all patients, a nasogastric tube was inserted in order to minimize the risk of pulmonary aspiration or of gastric distension.
For both groups of patients, ventilatory treatment was performed continuously for at least 4 to 6 h. This period could be lengthened, depending on clinical response and patient tolerance. Ventilatory treatment then was performed intermittently with sessions lasting 2 to 6 h tid. The duration of each session was determined by the improvement of the arterial blood gas levels and pH values, patient compliance, and the degree of ventilatory autonomy of the patient at the end of the session. The overall duration of noninvasive ventilation was established by the attending physician on the basis of clinical criteria and arterial blood gas levels.
During NPV and NIPPV, ECG activity, systemic BP, and breathing rate were regularly monitored in all patients. Arterial blood samples also were taken at regular intervals during the ventilatory session. A 30-min sampling frequency was employed during the first cycle of ventilation, and thereafter this frequency was reduced to at least one blood sample per ventilatory session.
The noninvasive ventilatory treatment was judged to be inadequate whenever there was evidence of an insufficient airway control or when one of the following conditions occurred: (1) the impossibility of obtaining satisfactory ventilation (ie, tidal volume, < 5 mL/kg) and a substantial improvement in gas exchange (ie, PaCO2 decrease, < 10 mm Hg; PaO2, > 60 mm Hg within 1 h of the start of mechanical ventilation) despite the optimum setting of the ventilator; and (2) a worsening of the comatose state despite an improvement in gas exchange (within 12 to 24 h of the start of mechanical ventilation). Under these circumstances, EI and conventional mechanical ventilation were performed. EI was not performed when patients or relatives (when patient was unconscious) refused this procedure.
Definition of the Study End Points
To assess the effectiveness of both ventilatory techniques in
each individual patient, we defined the following as the primary end
points: (1) death during RIICU stay; and (2) the need for intubation
due to the ineffectiveness of the noninvasive techniques. Treatment
failure was defined as death and or the need for EI for patients in
both groups. The need for EI was established by the attending
physicians on the basis of clinical worsening and the deterioration of
blood gas levels.
Secondary end points were the following: (1) complications (ie, pneumonia and pneumothorax) linked to mechanical ventilation; (2) duration of ventilatory assistance; and (3) length of stay at the RIICU.
Statistical Analysis
The normal distribution of variables was assessed by measuring
the skewness value. A skewness value > 1 indicated a distribution
that differed significantly from a normal distribution. Continuous
variables were compared using Student’s t test for normally
distributed variables and the Wilcoxon rank sum test for non-normally
distributed variables. The McNemar test with continuity correction was
used to compare the mortality and the treatment rates between the two
groups.11
The odds ratio of mortality and the treatment
failure rate with the relative 95% confidence interval (CI) was
determined using the Mantel-Haenszel test.11
A p value
< 0.05 was considered to be statistically significant. Unless
otherwise indicated, all data are presented as the mean ± SD.
|
Results |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Effectiveness of Matching
In the matched groups, the following levels of matching were
achieved: (1) 98.1% of patients were age-matched to within 5 years;
(2) 100% of patients were sex-matched; (3) 98.1% of patients had the
same causes triggering ARF; (4) 98.1% of patients had APACHE II scores
within 5 points of each other; (5) 98.1% of patients had pH values on
hospital admission that were within 0.03 of each other; and (6) 98.1%
of patients had PaCO2 values within
the defined criteria (Table 1
). The overall effectiveness of matching for the variables used reached
98.4%.
|
|
|
|
2 = 0.0002; difference was not
significant).The estimated mortality risk ratio was 0.857 (95% CI,
0.24 to 2.98).
|
Among the nine patients of the NIPPV group who died, three had been intubated and had died during conventional mechanical ventilation (multiple organ failure, one patient; acute myocardial infarction, one patient; sudden death, one patient), three patients had refused EI and had died during NIPPV (sepsis, two patients; cardiac arrest, one patient), two patients, although weaned from mechanical ventilation, had died of cardiac arrest and acute myocardial infarction, and the remaining patient had died of cardiac arrest during the period of weaning from NIPPV.
Treatment Failure:
The treatment failure rate was 20.7% in
patients of the NPV group and 24.5% in those of the NIPPV group (Table 5)
. Thirty-nine pairs of patients had concordant outcomes. Among the 14
discordant pairs, the McNemar test showed no statistical difference
(2 = 0.068; difference was not significant).The
estimated treatment failure risk ratio was 0.75 (95% CI, 0.21 to
2.49).
Secondary End Points
Complications: Five patients in the NPV group (9.4%) reported or developed side effects such as claustrophobia (5.7%), back pain (1.8%), and vomiting (1.8%) due to undergoing NPV.
Two patients in the NIPPV group (3.8%) developed pneumonia during noninvasive mechanical ventilation. Four patients in the NIPPV group (7.5%) reported skin abrasions caused by the employment of NIPPV.
Duration of Mechanical Ventilation: The total number of hours of mechanical ventilation was lower in patients in the NPV group (median, 24 h; range, 4 to 138 h) than in patients in the NIPPV group (median, 57 h; range, 24 to 172 h). The difference in the 38 concordant surviving pairs was statistically significant (p < 0.0001).
Hospital Stay: The median length of stay in the RIICUs for patients in the NPV group who survived was 11 days (range, 2 to 22 days), while the median length of stay of patients in the NIPPV group was 14 days (range, 8 to 31 days). The statistical analysis for the 38 matched pairs who survived showed a significant difference (p < 0.0001).
|
Discussion |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This retrospective study is the first which formally compares two different noninvasive ventilatory techniques for the treatment of COPD patients experiencing ARF. A crucial factor for the validity of this study is the success in matching patients of the NPV group with those of the NIPPV group for important confounding variables, especially the severity of ARF. The matching of patients in the NPV group with those in the NIPPV group was carried out using the following five variables that were significantly related to outcome in COPD patients admitted to an RIICU for ARF: age12 13 ; cause of exacerbation14 15 ; pH at RIICU admission16 17 ; PaCO2 at RIICU admission18 19 ; and APACHE II score.20 Moreover, as some studies have reported that sex affects long-term survival in COPD patients,21 22 we ensured that patients in the NPV group and the NIPPV group were sex-matched. The overall effectiveness of matching for these variables reached 98.4%. To verify the adequacy of the matching for the severity of chronic disease and ARF, we compared the patients in the NPV group and the NIPPV group using 11 other potentially confounding variables. Among these variables, no significant differences were observed between patients in the NPV group and those in the NIPPV group in terms of Glasgow coma scale score and albumin level at RIICU admission, HCO3 and pH level at discharge from the RIICU, the percentage of patients receiving home mechanical ventilation, the percentage of patients who had experienced previous episodes of ARF, and the percentage of patients with comorbidities. However, among the confounding variables, the PaO2/FIO2 ratio at RIICU admission was significantly lower in patients in the NPV group than in patients in the NIPPV group, and the number of patients requiring LTOT who had cor pulmonale was significantly higher in patients in the NPV group than in those in the NIPPV group. Furthermore, the degree of bronchial obstruction, analyzed in a subgroup of patients, was more severe in patients in the NPV group than in those in the NIPPV group. Because it has been reported23 that in COPD patients a low FEV1 is a negative prognostic factor for the outcome of acute decompensation, this finding may be considered an aggravating prognostic factor in patients in the NPV group. Furthermore, the data on the PaO2/FIO2 ratio at hospital admission and the number of patients requiring LTOT who had cor pulmonale suggest that acute decompensation and chronic disease could be more severe in patients in the NPV group than in those in the NIPPV group.
In the largest controlled study of NIPPV in acute COPD1 a treatment failure rate of 26% was reported in the study group. This rate is similar to the one we obtained in patients treated with NIPPV in the present study. The mortality rate was 9.3% in the study of Brochard and coworkers1 and 16.9% in our NIPPV group. Although comparisons between different studies exposes the results to bias, which must be taken into due account, it is important to stress that some outcome-related characteristics of patients at hospital admission, namely, age, pH, and PaO2/FIO2 ratio, were quite similar in the two studies.
It has been hypothesized that NPV may be superior to NIPPV in patients with more severe COPD7 and that NPV may be used as a rescue technique when patients do not respond to NIPPV before EI.8 Our data, showing that NPV has a similar treatment failure rate to that of NIPPV despite the higher degree of severity of some variables that are significantly related to outcome (ie, FEV1 level and presence of cor pulmonale),18 23 could be in keeping with the above hypothesis even though this point requires randomized controlled studies in order to be confirmed.
Another point resulting from the present study is that the number of hours of mechanical ventilation received and the length of stay in the RIICUs were significantly reduced in patients treated with NPV compared to those treated with NIPPV. However, these data need further evaluation, given that our study was retrospective and that no standardized criteria for the discontinuation of mechanical ventilation and hospital discharge were employed.
There are no published studies comparing the nurse workload during NPV and mask ventilation. However, we measured the time spent by nurses on the second day after admission to the hospital among COPD patients with ARF that was treated with use of an iron lung (unpublished data). The following procedures were considered: (1) transfer of the patient from the bed to inside the iron lung three times (total time, 105 min [this time also includes the placement of electrodes and monitor probes]); (2) measurement of tidal volume and minute ventilation by means of a Wright ventilograph (total time, 30 min); (3) tracheobronchial suction of secretion (total time, 35 min); (4) taking of arterial blood gas samples (total time, 30 min); (5) drug administration (total time, 20 min); and (6) other procedures (total time, 30 min). The total time spent by nurses for these procedures was 250 min. Nava et al24 have recently reported that nurse workload during mask ventilation in COPD patients was 540 min in the first 48 h and circa 200 min on the second day after admission.
NPV is actually used less often than mask ventilation. A recent Italian survey25 reported that an iron lung was used in 12% of patients with acute on chronic respiratory failure that had been treated with noninvasive mechanical ventilation. NIPPV is not without problems, and failure rates of 7 to 50% have been reported.26 The availability of both techniques could widen the field of application of noninvasive mechanical ventilation.
The present findings must be analyzed with caution by taking the following limitations into account: the retrospective design; the different teams involved in the management of these particular patients, which could imply different levels of expertise and a possible difference in the medical treatment employed; the lack of a protocol for the duration of noninvasive ventilation; and the lack of an objective protocol to set frequency and TI/TTOT ratio during noninvasive ventilation. All these factors may have influenced the results. It must be stressed, however, that the study was carried out in settings where the level of assistance was ensured by well-trained personnel with a high level of expertise in noninvasive mechanical ventilation and was performed during the same period.
In conclusion, the data of this preliminary study suggest that NPV is as effective as NIPPV in preventing EI during ARF in COPD patients. These findings need to be confirmed by a prospective, randomized, controlled trial comparing the two noninvasive ventilatory techniques.
|
Footnotes |
|---|
Received for publication January 19, 2001. Accepted for publication August 6, 2001.
|
References |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  G. G. Diaz, A. C. Alcaraz, J. C. P. Talavera, P. J. Perez, A. E. Rodriguez, F. G. Cordoba, and N. S. Hill Noninvasive Positive-Pressure Ventilation To Treat Hypercapnic Coma Secondary to Respiratory Failure Chest, March 1, 2005; 127(3): 952 - 960. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J Henderson Respiratory support of infants with bronchiolitis related apnoea: is there a role for negative pressure? Arch. Dis. Child., March 1, 2005; 90(3): 224 - 225. [Full Text] [PDF]
|
|
|
|
|
|
T. Todisco, S. Baglioni, A. Eslami, E. Scoscia, C. Todisco, L. Bruni, and M. Dottorini Treatment of Acute Exacerbations of Chronic Respiratory Failure: Integrated Use of Negative Pressure Ventilation and Noninvasive Positive Pressure Ventilation Chest, June 1, 2004; 125(6): 2217 - 2223. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Corrado, R. Ginanni, G. Villella, M. Gorini, A. Augustynen, D. Tozzi, A. Peris, S. Grifoni, A. Messori, C. Nozzoli, et al. Iron lung versus conventional mechanical ventilation in acute exacerbation of COPD Eur. Respir. J., March 1, 2004; 23(3): 419 - 424. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. Brochard Negative-Pressure Ventilation JAMA, February 26, 2003; 289(8): 983 - 983. [Full Text] [PDF]
|
|
|
|
|
|
A. Corrado and M. Gorini Negative-pressure ventilation: is there still a role? Eur. Respir. J., July 1, 2002; 20(1): 187 - 197. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
