- Email: [email protected]
Manual Hyperinflation Improves Alveolar Recruitment in Difficult-to-Wean Patients
Manual Hyperinflation Improves Alveolar Recruitment in Difficult-toWean Patients* Suh-Hwa Maa, DSN; Tzong-Jen Hung, MD; Kuang-Hung Hsu, PhD; Ya-I Hsieh, MS; Kwua-Yun Wang, MS; Chun-Hua Wang, MD; and Horng-Chyuan Lin, MD
Study objectives: To investigate the effect of manual hyperinflation (MH) in patients with atelectasis associated with ventilation support. Design: Patients were randomized to either an experimental group or a control group. Setting: Pulmonary ICUs from two hospitals. Patients: Twenty-three patients with atelectasis associated with ventilation support. Interventions: The MH technique was at a rate of 8 to 13 breaths/min for a period of 20 min each session, three times per day for 5 days. The control group received their standard prescribed mechanical ventilation without supplemental MH. Sputum contents (wet/dry weight ratio, viscosity), respiratory system capacity (spontaneous tidal volume [VT], maximal inspiratory pressure, rapid shallow breathing index [f/VT], chest radiograph signs, and PaO2/fraction of inspired oxygen [FIO2]) were measured just prior to the MH at day 0 as baseline, and at day 3 and day 6 of the study. Measurements and results: There were significant improvements in scores over the 6-day study in the experimental group compared to the control group in spontaneous VT (p ⴝ 0.035) and chest radiograph signs (p ⴝ 0.040), and a trend toward improvement of f/VT (p ⴝ 0.066) and PaO2/FIO2 (p ⴝ 0.061) after adjustment for covariates. Other outcome variables did not differ significantly between the experimental and control groups. Conclusions: MH performed on patients with atelectasis from ventilation support significantly improved alveolar recruitment. (CHEST 2005; 128:2714 –2721) Key words: alveolar recruitment; atelectasis; difficult to wean; manual hyperinflation Abbreviations: ANOVA ⫽ analysis of variance; Fio2 ⫽ fraction of inspired oxygen; f/Vt ⫽ rapid shallow breathing index; MH ⫽ manual hyperinflation; OR ⫽ odds ratio; Pimax ⫽ maximal inspiratory pressure; Vt ⫽ tidal volume
ventilation is indicated in acute reM echanical versible respiratory failure. However, patients receiving mechanical ventilation may have an increased risk of sputum retention, atelectasis, and
*From the School of Nursing (Dr. Maa), Department of Business Administration (Dr. Hsu), Department and Graduate Institute of Health Care Management, and Department of Thoracic Medicine II (Drs. C-H Wang and Lin), Chang Gung University, Tao-Yuan; Department of Thoracic Medicine (Dr. Hung), Wei Gong Memorial Hospital, Miao-Li; Department of Nursing (Ms. Hsieh), Taipei Veterans General Hospital, Taipei; and School of Nursing (Ms. K-Y Wang), National Defense Medical Center, Taipei, Taiwan. This study was supported by the National Science Council of Taiwan, contract No. NSC 90 –2314-B-182– 062. Manuscript received February 2, 2005; revision accepted May 3, 2005. Reproduction of this article is prohibited without written permission from the American College of Chest Physicians (www.chestjournal. org/misc/reprints.shtml). Correspondence to: Suh-Hwa Maa, DSN, School of Nursing, Chang Gung University, 259, Wen-Hwa First Rd, Kwei-San, Tao-Yuan, Taiwan, ROC; e-mail: [email protected] 2714
pneumonia,1,2 making ventilation weaning more difficult3 and resulting in excess morbidity and mortality. The cost of maintaining patients on prolonged ventilation in the ICUs of acute care hospitals are high.4 Thus, every effort should be made to determine which patients can be rapidly extubated so as to keep the weaning period to a minimum. Previous evidence suggests that manual hyperinflation (MH) can mobilize pulmonary secretions, reverse atelectatic alveoli, and increase alveolar oxygenation. Many clinical studies have reported the short-term benefits of MH on sputum clearance,3,5 reexpansion of atelectasis,5– 8 improvement of dynamic compliance,9 and oxygenation.10 –12 However, the lack of standardized methods for the delivery of MH makes the synthesis and interpretation of the findings difficult. The variability in the types of MH circuit (selfinflating manual resuscitation bags vs oxygen-powered, manual resuscitation bags), the method of MH Clinical Investigations in Critical Care
delivery (technique of pausing at full inspiration vs pressing half of the resuscitator), and the variability in the dosage of MH (ie, duration) all point toward the need for further development of the knowledge base in order to guide best practice. Manual hyperinflation is defined as inflating the lungs using oxygen and manual compression to provide a tidal volume (Vt) exceeding baseline Vt, and using a Vt that is 50% greater than that delivered by the ventilator, requiring a peak inspiratory pressure of from 20 to 40 cm H2O.13 Four factors are considered important in performing the MH technique: the application of larger-than-normal Vt breaths,14,15 use of a slow inspiratory flow rate,16,17 an inspiratory pause,18,19 and a pressure manometer.20,21 In addition, the quick release of pressure on expiration leading to a rapid flow of air can simulate the effect of a cough.11,22,23 Even though no comprehensive studies have been done that incorporate and evaluate all four of the important MH techniques, there is support from clinical and research literature on the theoretical foundations and effectiveness of each of the factors separately. First, the use of larger-than-normal Vt is based on the hypothesis that by delivering a largervolume breath over time, MH may increase the expiratory flow rate and assist in moving secretions toward more proximal airways, where they can be cleared by suctioning.24 Second, the rate of inflation of the lung as a whole is a function of inflation pressure, compliance, and airway resistance. Nunn25 described the response to passive inflation of the lungs by the development of a constant airway pressure. If a constant inflation pressure is maintained, an alveolus with half the compliance but twice the resistance of another alveolus will increase in volume by half the volume change of the other alveolus. Thus, the relative distribution of gas between the two alveoli is independent of the rate or duration of inflation. In addition, using both hands to compress the bag can produce a Vt that is 50% greater than that delivered by the ventilator.26 Furthermore, the rate at which the bag is compressed, rather than the resistance of the circuit itself, is the main influence on the peak inspiratory flow rate.16,17,27 A fast inflation rate that does not allow the reservoir bag to fill adequately, and reduces the fraction of the inspired oxygen (Fio2).28,29 Moreover, delivering an increased Vt via MH may generate adequate transpulmonary pressure gradients to overcome alveolar atelectasis. Atelectatic alveoli do not reexpand immediately when the ventilator cycles with the inspiratory phase because a variable period of time is required before the alveolar critical opening pressure is reached.30 Therefore, the third important factor, the use of an inspiratory hold during www.chestjournal.org
MH, is thought to maintain these pressure gradients for an appropriate length of time. This technique may influence the distribution of the ventilation25 and allow time for alveolar inflation or enlargement, as well as the recruitment or unfolding of interdependent atelectatic alveoli. Finally, a pressure manometer can improve the performance of MH and optimize both the safety and the effectiveness of the treatment.31 Although there is no consensus about specific safe upper limits for peak airway pressure, barotraumas manifest at peak airway pressures of 26 to 64 cm H2O as demonstrated by several animal studies.20 Therefore, it is reasonable and prudent to minimize the peak airway pressure as much as possible during MH or any other ventilatory support procedure.32 This study examines the effect of MH in patients with atelectasis associated with ventilation support. The foundation for the practice of MH in this study is based on the best evidence available from clinical and research literature, and incorporates all four factors of MH technique considered to be important, as described above. The hypothesis of this study is that those in the experimental group should have improved sputum contents (wet/dry weight ratio, viscosity), respiratory system capacity (spontaneous Vt), maximal inspiratory pressure (Pimax), an improved rapid shallow breathing index (f/Vt), as well as improved chest radiograph signs and oxygenation ratio (Pao2/Fio2).
Materials and Methods Patient Selection Thirty-three patients with atelectasis due to ventilatory support were recruited at the pulmonary ICUs from two hospitals; of these 33 patients, 23 completed all of the study procedures. This study was a two-group, prospective, randomized study lasting 6 days. Patients were assigned to one of two groups: standard care with supplemental MH (experimental group, n ⫽ 10) or standard care only (control group, n ⫽ 13). The Institutional Ethical Committee approved the study protocol, and all patients gave informed written consent. The inclusion criteria consisted of the following: age ⱖ 40 years; ventilation support ⬎ 7 days and a positive end-expiratory pressure from 6 to 8 cm H2O; pulmonary atelectasis; excessive secretions (⬎ 30 mL/d); and spontaneous Vt ⬍ 250 mL and/or Pimax ⬍ 25 cm H2O and/or Vt ⬍ 400 mL under ventilator assistance. Pulmonary atelectasis was diagnosed using the following: (1) chest radiography showing increased infiltration, and (2) physical examination revealing weakness or muteness of the sounds in the involved area. The structural changes that develop in atelectasis increase the density of the lungs. The increase in lung density resists radiograph penetration and is revealed on radiograph films as increased opacity (ie, whiter in appearance). Thus, the more severe the atelectasis, the denser the lungs, and the whiter the radiograph film. Patients were assessed clinically and with a chest radiograph at recruitment to ensure the absence of a Fio2 ⱖ 0.6 requirement, CHEST / 128 / 4 / OCTOBER, 2005
2715
pulmonary pathology (for example, ARDS), active infection, acute cardiovascular dysfunction, or other systemic diseases. The trial took place between January 2001 and June 2001. There were three different types of ventilation systems used throughout this study (models 7200, 740, or 760; Nellcor Puritan Bennett; Temecula, CA). Standard Care All subjects were asked to continue any current prescribed medication (such as anticholinergic inhaled agents, inhaled corticosteroids, theophylline, prednisolone, or erythromycin) and chest physiotherapy (such as chest percussion, positioning, and suction) throughout the experiment. For the control group, these were the only prescribed treatments. None of the subjects received any sedation or narcotics. MH To ensure that uniform and correct techniques were employed, MH was administered by only one investigator. A 2.0-L reusable manual resuscitator (model 2153 MR100 plus; Galemed Corporation; Taipei, Taiwan) was used to deliver the MH breaths, and was connected to a flow of 100% oxygen at 15 L/min (calibrated with an oxygen analyzer). A force meter (Inspiratory Force Meter; Boehringer Laboratories; Norristown, PA) was connected between the resuscitator and the patient. Patients received MH to a peak airway pressure of 20 cm H2O by use of the resuscitator. The resuscitator was slowly compressed with both hands, and an inspiratory breath was maintained for 3 to 5 s at the end of pressing half of the resuscitator, and then completely pressing the resuscitator. Expiration was passive and unobstructed to facilitate expiratory flow with no positive end-expiratory pressure applied. Sufficient time was allowed for the resuscitator to fill completely prior to the next breath. Airway suctioning of the endotracheal tube was performed using size 14 catheters (Pahsco; Pacific Hospital Supply; Taipei, Taiwan) at the end of the MH procedure. The MH procedure was carried out at a rate of 8 to 13 breaths/min for a period of 20 min for each session tid (at 7:40 am, 11:40 am, and 3:40 pm) for 5 days on days 1 to 5 of the study. Sputum Sampling Nurses were instructed to collect and record the total amount of daily sputum (milliliters per 24 h) throughout the study. An aliquot of sputum from each patient’s total amount of daily sputum was freeze-dried (at – 80°C, at a negative pressure of 40 cm H2O) overnight to measure the wet/dry weight ratio. The viscosity of the sputum was measured using a viscometer at room temperature (25°C) with distilled water as a control, using sputum sampled by nurses at 7 am on days 0 (baseline), 3, and 6 of the study. After receiving chest percussion, the sputum was collected by airway suctioning of the endotracheal tube into a sterile pot. Measurement of Respiratory System Capacity Respiratory system capacity measurements were obtained 30 min after sputum sampling and just prior to the 7:40 am MH on days 0 (baseline), 3, and 6 of the study. The spontaneous Vt score was measured during ventilation disconnection (Haloscale Wright Respirometer; Ferraris Medical Limited; Middlesex, UK) by a respiratory therapist as spontaneous respiratory volume (milliliters) per minute divided by respiratory rate per minute. The Pimax was measured during ventilation disconnection with the inspiratory force meter by the respiratory therapist. A 2716
unidirectional expiratory valve pressure-manometer was connected to the endotracheal tube or tracheostomy, the port was occluded at end-expiration for 20 s, and after three spontaneous maximal inspiratory efforts the Pimax was recorded.33,34 The f/Vt score was measured during mechanical ventilation as a calculation of the ratio of the respiratory rate per minute (frequency) to the Vt setting (liters) from the display on the ventilation. A portable radiograph machine was used, and a staff radiologist reported the chest radiograph findings each morning. Scores were given as 1 (improved) or 2 (not improved). Nurses also recorded the ventilator volume and measured the cuff pressure (Control-inflator; VBM Medizintechnik; Sulz am Neckar, Germany) at least once or twice daily for patients with tracheostomy. The volume needed to attain a full seal should be recorded at least once or twice daily. The need for increasingly larger volumes indicates an expanding trachea. The pressure was kept at levels ⬍ 20 mm Hg. If there was an air leak in the cuff or cuff inflation system, nurses reinflated the cuff via a stopcock. If the ventilator could be set to compensate for the leak, the patient was not reintubated. If significant aspiration or inadequate ventilation was present, a new tube was inserted. Oxygenation Ratio Pao2/Fio2 was measured during mechanical ventilation as derived from arterial blood gas analysis and the Fio2 on days 0 (baseline) and 6 of the study. A calibrated blood gas analyzer (model 278; Ciba-Corning; Medfield, MA) was used for arterial blood gas analysis, and the Fio2 was read from the display on the ventilator obtained just prior to MH. Statistical Analysis Statistical software (version 10.0; SPSS; Chicago, IL) was used for data analysis. 2 test and Fisher Exact Test were used to assess the success of the randomization process in achieving two comparable groups. A t test was performed to establish the baseline stability of the dependent variables. A repeated-measures analysis of variance (ANOVA) was performed to compare scores over time between the experimental and the control groups on each of the seven dependent variables measured at each of the three time points: day 0 (baseline), day 3, and day 6 of the study. This method accounted for six covariates: sex (male, female), age, setting (medical center, local hospital), intubation (endotracheotomy, endointubation), logarithm of length of ventilation prior to enrolment, and logarithm of total sputum amount.35 The null hypothesis is that there is no interaction between group and study duration, ie, the 5 days of treatment with repeated measures on days 0, 3, and 6. There is a gradual increase in treatment effect if there is interaction between the group and the duration of the treatment. The Mantel-Haenszel 2 test for categorical data, adjusted odds ratio (OR), and multiple logistic regression were performed on the chest radiograph scores. The OR was calculated as the odds in favor of clinical improvement in the treatment group divided by the odds in favor of clinical improvement in the control group. Significance was indicated at p ⬍ 0.05.
Results Sample Initially, 33 patients agreed to participate in this investigation; of these, 10 patients withdrew. Twenty-three patients (n ⫽ 23) completed the full course Clinical Investigations in Critical Care
of treatment. The attrition rate in this experiment was high (30%), perhaps because of the characteristics of respiratory failure from mechanical ventilator support. Of the 10 patients who dropped out, 3 died (1 from the experimental group and 2 from the control group), 4 withdrew voluntarily (3 from the experimental group and 1 from the control group), and 3 received oxygen content ⬎ 50% during the study period (1 from the experimental group and 2 from the control group). As most of the withdrawals were for medical reasons, there is reason to hypothesize that some of these patients might have benefited from MH. In addition, there were no statistically significant differences in baseline demographics, clinical characteristics, or outcome measurements between those who withdrew and the remaining participants. Patient Characteristics Table 1 lists the sample baseline demographic and clinical characteristics including intubation, setting, and length of ventilation prior to enrollment. There were no statistically significant differences between the subjects of the two groups. The sample contained more men (n ⫽ 17) than women (n ⫽ 6); 74% of the subjects were ⬎ 65 years of age, 100% were married, and 74% had no history of smoking. All subjects were receiving mechanical ventilation for at least 7 days prior to study entry. The average Fio2 was 35%; pneumonia was diagnosed in 17 persons, and lowerlobe atelectasis was found in 19 persons.
Table 1—Characteristics of Subjects* Characteristics
Experimental Control All p Group Group Subjects Value†
Total subjects, No. 10 Sex Male 8 (80) Female 2 (20) Age, yr ⱖ 65 7 (70) ⱕ 64 3 (30) Cigarette history No 6 (60) Yes 4 (40) Intubation Tracheostomy 3 (30) Endotracheal tube 7 (70) Setting Medical center 8 (80) Local hospital 2 (20) Length of ventilation prior to enrollment, d 7 5 (50) 8 to 13 2 (20) ⱖ 14 3 (30)
13
Outcome Measures Outcome measure scores are compared in Table 3, and the mean scores and the p value for their group ⫻ time interaction in repeated-measures ANOVAs adjusted for covariates are listed. The spontaneous Vt and chest radiograph scores show significant differences between the experimental and control groups, and the f/Vt and Pao2/Fio2 scores show a trend toward improvement in the experimental group compared to the control group. In the experimental group, spontaneous Vt scores of 196.3 mL at baseline increased to 270.5 mL on day 6 (indicating an improvement) compared to the control group, which increased from 208.49 mL at baseline to 220.14 mL on day 6 (p ⫽ 0.035; Fig 1). Furthermore, in the experimental group, f/Vt scores of 216.59 at baseline decreased to 150.21 on day 6 compared to the control group, which decreased from 174.04 to 164.74 (p ⫽ 0.066; Fig 2). Moreover, in the experimental group, Pao2/Fio2 scores increased from 222.07 at baseline to 264.45 on day 6 compared to the control group, which decreased from 228.64 to 203.53 (p ⫽ 0.061; Fig 3). Further, in the experimental group, chest radiograph scores improved 15.55-fold (95% confidence interval, 1.14 to 239.77; p ⫽ 0.040) after adjustment for covariates
23 0.46
9 (69) 4 (31)
17 (74) 6 (26) 0.54
10 (77) 3 (23)
0.20 11 (85) 2 (15)
17 (74) 6 (26)
5 (39) 8 (61)
8 (35) 15 (65)
0.51
0.31 8 (61) 5 (39)
16 (70) 7 (30) 0.40
3 (23) 3 (23) 7 (54)
Table 2—Comparison of Baseline Measurements Between Groups (n ⴝ 23)*
17 (74) 6 (26)
8 (35) 5 (22) 10 (43)
*Data are presented as No. (%) unless otherwise indicated. †2 and Fisher Exact Test. www.chestjournal.org
Table 2 shows the actual values of the outcome variables in all patients at baseline. Differences in baseline values between the two groups were not statistically significant except for the sputum amount. The logarithm of the total sputum amount was used as covariates in the repeated-measures ANOVA model; therefore, the data from all the subjects were included in the analyses.
Variables Sputum content Amount, mL/24 h Wet/dry weight ratio, % Viscosity, min Respiratory system capacity Spontaneous Vt, mL Pimax, cm H2O f/Vt Oxygenation ratio Pao2/Fio2
Experimental Group (n ⫽ 10)
Control p Group (n ⫽ 13) Value†
186.50 ⫾ 116.57 27.82 ⫾ 11.48
94.62 ⫾ 53.64 19.29 ⫾ 13.67
0.02 0.13
9.67 ⫾ 17.91
8.57 ⫾ 12.60
0.87
196.30 ⫾ 80.87 208.49 ⫾ 54.56 27.00 ⫾ 16.87 20.92 ⫾ 10.46 216.59 ⫾ 146.62 174.04 ⫾ 66.62
0.67 0.30 0.36
222.07 ⫾ 93.94
0.90
228.64 ⫾ 131.84
*Data are presented as mean ⫾ SD. †t test. CHEST / 128 / 4 / OCTOBER, 2005
2717
Table 3—Repeated-Measures ANOVA on Outcome Measures Between Experimental and Control Groups (n ⴝ 23)* Outcome Measures Sputum content Sputum wet/dry ratio, % Day 0 (baseline) Day 1 Day 2 Day 3 Day 4 Day 5 Sputum viscosity, min Day 0 (baseline) Day 3 Day 6 Respiratory system capacity Spontaneous Vt, mL Day 0 (baseline) Day 3 Day 6 Pimax, cm H2O Day 0 (baseline) Day 3 Day 6 f/Vt Day 0 (baseline) Day 3 Day 6 Chest radiographs Improved/not improved, No. Adjusted OR 95% confidence interval for OR Oxygenation ratio Pao2/Fio2 Day 0 (baseline) Day 6
Experimental Group (n ⫽ 10)
Control Group (n ⫽ 13)
27.82 ⫾ 11.48 28.36 ⫾ 11.32 32.01 ⫾ 19.08 25.83 ⫾ 9.25 48.76 ⫾ 51.59 37.22 ⫾ 35.99
19.29 ⫾ 13.67 19.38 ⫾ 12.65 20.52 ⫾ 13.95 20.27 ⫾ 18.00 19.89 ⫾ 16.07 18.87 ⫾ 14.28
9.67 ⫾ 17.91 5.85 ⫾ 4.98 5.49 ⫾ 5.35
8.57 ⫾ 12.60 13.01 ⫾ 21.08 24.60 ⫾ 23.95
196.30 ⫾ 80.87 287.07 ⫾ 120.03 270.50 ⫾ 98.65
208.49 ⫾ 54.56 223.61 ⫾ 63.26 220.14 ⫾ 79.34
27.00 ⫾ 16.87 30.30 ⫾ 10.26 36.10 ⫾ 16.16
20.92 ⫾ 10.46 17.92 ⫾ 9.54 18.38 ⫾ 8.14
216.59 ⫾ 146.62 133.67 ⫾ 84.08 150.21 ⫾ 66.12
174.04 ⫾ 66.62 144.22 ⫾ 65.55 164.74 ⫾ 101.50
9/1 16.56 1.14–239.77
6/7 1.00
222.07 ⫾ 93.94 264.45 ⫾ 113.41
228.64 ⫾ 131.84 203.53 ⫾ 96.17
p Value† 0.831
0.145
0.035
0.194
0.066
0.040
0.061
*Data are presented as mean ⫾ SD unless otherwise indicated. †Treatment ⫻ time interaction in repeated-measures ANOVA, adjusted for sex, age, setting, intubation, logarithm of length of ventilation prior to enrollment, and logarithm of total sputum amount.
Figure 1. Mean values of spontaneous Vt in both groups; horizontal bars ⫽ ⫹ 1 SD. p ⫽ 0.035 refers to differences between groups over time, with changes only on day 3 and day 6 while adjusting for sex, age, setting, intubation, logarithm of length of ventilation prior to enrollment, and logarithm of total sputum amount. Higher spontaneous Vt values represent patients with atelectasis associated with ventilation support-improved alveolar recruitment. 2718
Figure 2. Mean values of f/Vt in both groups; horizontal bars ⫽ ⫹ 1 SD. p ⫽ 0.066 refers to differences between groups over time, with changes only on day 3 and day 6 while adjusting for sex, age, setting, intubation, logarithm of length of ventilation prior to enrollment, and logarithm of total sputum amount. Lower f/Vt values represent patients with atelectasis associated with ventilation support-improved alveolar recruitment. Clinical Investigations in Critical Care
Figure 3. Mean values of Pao2/Fio2 in both groups; horizontal bars ⫽ ⫹ 1 SD. p ⫽ 0.061 refers to differences between groups over time, with changes only on day 6 while adjusting for sex, age, setting, intubation, logarithm of length of ventilation prior to enrolment, and logarithm of total sputum amount. Higher Pao2/ Fio2 values represent patients with atelectasis associated with ventilation support-improved alveolar recruitment.
compared with the control group. Scores in sputum wet/dry weight ratio increased from 27.82% at baseline to 37.42% on day 6 in the experimental group and decreased from 19.29 to 18.87% in the control group; scores in sputum viscosity decreased from 9.67 min at baseline to 5.49 min on day 6 in the experimental group, and increased from 8.57 to 24.60 min in the control group, Pimax scores of 27.00 cm H2O at baseline increased to 36.10 cm H2O on day 6 in the experimental group, and decreased from 20.92 to 18.38 cm H2O in the control group after adjustment for covariates compared with the control group, even though these changes were not statistically significant.
Discussion To our knowledge, this is the first study to examine the potential benefits of MH to 20 cm H2O by oxygen-powered, manual resuscitation bag with an inspiratory breath-hold of 3 to 5 s, while pressing half of the resuscitator, in a group of intubated patients with atelectasis. In addition, this study employed methodologic features that strengthened validity and reliability of the findings, including the randomization of subjects to groups and multivariate analysis controlling for known covariates. Those receiving MH had statistically significant www.chestjournal.org
improvement in respiratory system capacity and oxygenation ratio, evidenced by the scores of spontaneous Vt and chest radiograph signs, and a trend toward improvement of f/Vt and Pao2/Fio2 compared to the control groups, after adjusting for the effects of sex, age, setting, intubation, logarithm of length of ventilation prior to enrolment, and logarithm of total sputum amount. These results are not consistent with the fact that the MH technique was initially designed to enhance clearance of airway secretions.27 Nevertheless, it supported the initial hypothesis of this study, that MH improves alveolar recruitment by delivering a larger-volume breath over time,24 and by the development of a constant airway pressure25 in patients with atelectasis from mechanical ventilation. In addition, MH produced no adverse events in the experimental group, as none of the patients experienced pneumothorax, suffocation, or hypotension during or following MH. However, potential limitations of this investigation should be considered when interpreting the findings. These include the following: many of the outcome measures have a subjective component to them, because the respiratory therapist who scored these outcomes was not blinded; failure to obtain outcome data of successfully weaned patients before they completed this study; length of ventilation; and the small sample size. In addition, this model did not account for cigarette consumption, length of atelectasis, and the severity of the illness, all of which could account for group differences over time. Future studies should incorporate such risk adjustment using standard severity of illness measures as APACHE (acute physiology and chronic health evaluation)36 or sequential organ failure assessment37 scores. Further study on the effects of MH should be conducted using different resuscitation circuits (such as MH to 30 or 35 cm H2O), different subject groups (such as receiving ventilatory support for ⬍ 7 days, and differentiating between chronic and acute atelectasis), and different operators (such as physiotherapists, respiratory therapists, or nurses). Additional studies are also needed to elucidate the long-term outcomes such as time to extubation, time to discharge, ventilator-free days, and discharge status (home, long-term pulmonary care, death). Specifications by the practitioner, and patient preferences for treatment duration and frequency should also be explored.
Conclusion This study provides evidence that MH performed in a stable patient with atelectasis associated with CHEST / 128 / 4 / OCTOBER, 2005
2719
ventilation can improve spontaneous Vt and chest radiograph signs, and a trend toward improvement of f/Vt and Pao2/Fio2. MH is a nursing intervention that could be implemented without a physician’s order, and has the potential to make a positive impact on patient outcome. While this preliminary study with a small sample size does not warrant changes in clinical practice at this time, it does contribute to the evidence base on the benefits of MH in critically ill and ventilator-dependent patients. Further investigations are required to replicate this study with a larger sample size, evaluate different techniques of MH, evaluate effects of MH in different patient groups, and determine the longterm outcomes of MH. As we continue to build evidence through additional studies, we may eventually be able to recommend practice guidelines for the procedure of MH for health professionals to treat various clinical conditions. ACKNOWLEDGMENT: The authors like to express their appreciation to Ivo L. Abraham, PhD, RN, of Matrix45, LLC, and the School of Nursing, University of Pennsylvania, Philadelphia, PA, and Karen M. MacDonald, PhD, RN, of Matrix45, LLC, Earlysville, VA, for the time and effort they spent commenting on earlier versions of this article. The authors also like to acknowledge the contribution of the physiotherapy, nursing, and medical staff of Chang Gung Memorial Hospital for their expert assistance, especially Dr. Han-Pin Kuo.
References 1 Konrad F, Schreiber T, Brecht-Kraus D, et al. Mucociliary transport in ICU patients. Chest 1994; 105:237–241 2 Anderson J, Jenkins S. Physiotherapy problems and their management. In: Webber B, Pryor J, ed. Physiotherapy for respiratory and cardiac problems. London, UK: Churchill Livingstone, 1993; 226 –227 3 Berney S, Denehy L. A comparison of the effects of manual and ventilator hyperinflation on static lung compliance and sputum production in intubated and ventilated intensive care patients. Physiother Res Intern 2002; 7:100 –108 4 Wagner DP. Economics of prolonged mechanical ventilation. Am Rev Respir Dis 1989; 140(suppl):14S–18S 5 Hodgson C, Denehy L, Ntoumenopoulos G, et al. An investigation of the early effects of manual lung hyperinflation in critically ill patients. Anesth Intensive Care 2000; 28:255–261 6 McCarren B, Chow CM. Description of manual hyperinflation in intubated patients with atelectasis. Physiother Theory Pract 1998; 14:199 –210 7 Rothen HU, Sporre B, Engberg G, et al. Re-expansion of atelectasis during general anesthesia: a computed tomography study. Br J Anaesth 1993; 71:788 –795 8 Stiller K, Geake T, Taylor J, et al. Acute lobar atelectasis: a comparison of two chest physiotherapy regimens. Chest 1990; 98:1336 –1340 9 Barker M, Adams S. An evaluation of a single chest physiotherapy treatment on mechanically ventilated patients with acute lung injury. Physiother Res Intern 2002; 7:157–169 10 Patman S, Jenkins S, Stiller K. Manual hyperinflation: effects on respiratory parameters. Physiother Res Intern 2000; 5:157–171 2720
11 Jones AYM, Hutchinson RC, Oh TE. Effects of bagging and percussion on total static compliance of the respiratory system. Physiotherapy 1992; 78:661– 666 12 Tweed WA, Phua WT, Chong KY, et al. Tidal volume, lung hyperinflation and arterial oxygenation during general anaesthesia. Anaesth Intensive Care 1993; 21:806 – 810 13 Denehy L. The use of manual hyperinflation in airway clearance. Eur Respir J 1999; 14:958 –965 14 Maxwell L, Ellis E. Secretion clearance by manual hyperinflation: possible mechanisms. Physiother Theory Pract 1998; 14:189 –197 15 McCarren B, Chow CM. Manual hyperinflation: a description of the technique. Aust J Physiother 1996; 42:203–208 16 Baker AB, Colliss JE, Cowie RW. Effects of varying inspiratory flow waveform and time in intermittent positive pressure ventilation: various physiological variables. Br J Anaesth 1977; 49:1221–1233 17 Pillet O, Choukroun ML, Castaing Y. Effects of inspiratory flow rate alterations on gas exchange during mechanical ventilation in normal lungs: efficiency of end-inspiratory pause. Chest 1993; 103:1161–1165 18 Bindslev L, Santesson J, Hedenstierna G. Distribution of inspired gas to each lung in anesthetized human subjects. Acta Anaesth Scand 1981; 25:297–302 19 Dammann J, McAslan T, Maffer C. Optimal flow pattern for mechanical ventilation of the lungs. Crit Care Med 1978; 6:293–310 20 Redfern J, Ellis E, Holmes W. The use of a pressure manometer enhances student physiotherapists’ performance during manual hyperinflation. Aust J Physiother 2001; 47: 121–131 21 Hila J, Ellis E, Holmes W. Feedback withdrawal and changing compliance during manual hyperinflation. Physiother Res Intern 2002; 7:53– 64 22 Hack I, Katz C, Eales C. Airway pressure changes during “bag squeezing.” S Afr J Philos 1980; 36:97–99 23 Windsor H, Harrison G, Nicholson T. “Bag squeezing”: a physiotherapeutic technique. Med J Aust 1972; 2:829 – 832 24 Clement AJ, Hubsch SK. Chest physiotherapy by the ‘bag squeezing’ method. Physiotherapy 1968; 54:355–359 25 Nunn JF. Nunn’s applied respiratory physiology. 5th ed. Oxford, UK: Butterworth-Heinemann, 2000; 166 26 Chulay M, Graeber GM. Efficacy of a hyperinflation and hyperoxygenation suctioning intervention. Heart Lung 1988; 17:15–22 27 Maxwell L. Ellis ER. The effects of three manual hyperinflation techniques on pattern of ventilation in a test lung model. Anaesth Intensive Care 2002; 30:283–288 28 Eaton JM. Adult manual resuscitators. Br J Anaesth 1984; 31:67–70 29 Glass C, Grap M, Corley M, et al. Nurses’ ability to achieve hyperinflation and hyperoxygenation with a manual resuscitation bag during endotracheal suctioning. Heart Lung 1993; 22:158 –165 30 Galvis AG, Bowen A, Oh KS. Nonexpandable lung after drainage of pneumothorax. Am J Respir Crit Care Med 1981; 136:1224 –1226 31 Goldstein B, Catlin EA, Vetere JM, et al. The role of in-line manometers in minimizing peak and mean airway pressure during hand regulated ventilation of newborn infants. Respir Care 1989; 34:23–27 32 Haake R, Schlichtig R, Ulstad D, et al. Barotrauma: pathophysiology, risk factors, and prevention. Chest 1987; 91:608 – 613 33 Marini JJ, Smith TC, Lamb VJ. Estimation of inspiratory muscle strength in mechanically ventilated patients: the Clinical Investigations in Critical Care
measurement of maximal inspiratory pressure. J Crit Care 1986; 1:32–38 34 Caruso P, Friedrich C, Denari SDC, et al. The unidirectional valve is the best method to determine maximal inspiratory pressure during weaning. Chest 1999; 115:1096 –1101 35 Tukey JW. Exploratory data analysis. Reading, MA: AddisonWesley, 1977
www.chestjournal.org
36 Knaus WA, Draper EA, Wagner DP, et al. APACHE II: a severity of disease classification system. Crit Care Med 1985; 13:818 – 829 37 Vincent JL, Moreno R, Takala J, et al. The SOFA score to describe organ dysfunction/failure. Intensive Care Med 1996; 22:707–710
CHEST / 128 / 4 / OCTOBER, 2005
2721
Study objectives: To investigate the effect of manual hyperinflation (MH) in patients with atelectasis associated with ventilation support. Design: Patients were randomized to either an experimental group or a control group. Setting: Pulmonary ICUs from two hospitals. Patients: Twenty-three patients with atelectasis associated with ventilation support. Interventions: The MH technique was at a rate of 8 to 13 breaths/min for a period of 20 min each session, three times per day for 5 days. The control group received their standard prescribed mechanical ventilation without supplemental MH. Sputum contents (wet/dry weight ratio, viscosity), respiratory system capacity (spontaneous tidal volume [VT], maximal inspiratory pressure, rapid shallow breathing index [f/VT], chest radiograph signs, and PaO2/fraction of inspired oxygen [FIO2]) were measured just prior to the MH at day 0 as baseline, and at day 3 and day 6 of the study. Measurements and results: There were significant improvements in scores over the 6-day study in the experimental group compared to the control group in spontaneous VT (p ⴝ 0.035) and chest radiograph signs (p ⴝ 0.040), and a trend toward improvement of f/VT (p ⴝ 0.066) and PaO2/FIO2 (p ⴝ 0.061) after adjustment for covariates. Other outcome variables did not differ significantly between the experimental and control groups. Conclusions: MH performed on patients with atelectasis from ventilation support significantly improved alveolar recruitment. (CHEST 2005; 128:2714 –2721) Key words: alveolar recruitment; atelectasis; difficult to wean; manual hyperinflation Abbreviations: ANOVA ⫽ analysis of variance; Fio2 ⫽ fraction of inspired oxygen; f/Vt ⫽ rapid shallow breathing index; MH ⫽ manual hyperinflation; OR ⫽ odds ratio; Pimax ⫽ maximal inspiratory pressure; Vt ⫽ tidal volume
ventilation is indicated in acute reM echanical versible respiratory failure. However, patients receiving mechanical ventilation may have an increased risk of sputum retention, atelectasis, and
*From the School of Nursing (Dr. Maa), Department of Business Administration (Dr. Hsu), Department and Graduate Institute of Health Care Management, and Department of Thoracic Medicine II (Drs. C-H Wang and Lin), Chang Gung University, Tao-Yuan; Department of Thoracic Medicine (Dr. Hung), Wei Gong Memorial Hospital, Miao-Li; Department of Nursing (Ms. Hsieh), Taipei Veterans General Hospital, Taipei; and School of Nursing (Ms. K-Y Wang), National Defense Medical Center, Taipei, Taiwan. This study was supported by the National Science Council of Taiwan, contract No. NSC 90 –2314-B-182– 062. Manuscript received February 2, 2005; revision accepted May 3, 2005. Reproduction of this article is prohibited without written permission from the American College of Chest Physicians (www.chestjournal. org/misc/reprints.shtml). Correspondence to: Suh-Hwa Maa, DSN, School of Nursing, Chang Gung University, 259, Wen-Hwa First Rd, Kwei-San, Tao-Yuan, Taiwan, ROC; e-mail: [email protected] 2714
pneumonia,1,2 making ventilation weaning more difficult3 and resulting in excess morbidity and mortality. The cost of maintaining patients on prolonged ventilation in the ICUs of acute care hospitals are high.4 Thus, every effort should be made to determine which patients can be rapidly extubated so as to keep the weaning period to a minimum. Previous evidence suggests that manual hyperinflation (MH) can mobilize pulmonary secretions, reverse atelectatic alveoli, and increase alveolar oxygenation. Many clinical studies have reported the short-term benefits of MH on sputum clearance,3,5 reexpansion of atelectasis,5– 8 improvement of dynamic compliance,9 and oxygenation.10 –12 However, the lack of standardized methods for the delivery of MH makes the synthesis and interpretation of the findings difficult. The variability in the types of MH circuit (selfinflating manual resuscitation bags vs oxygen-powered, manual resuscitation bags), the method of MH Clinical Investigations in Critical Care
delivery (technique of pausing at full inspiration vs pressing half of the resuscitator), and the variability in the dosage of MH (ie, duration) all point toward the need for further development of the knowledge base in order to guide best practice. Manual hyperinflation is defined as inflating the lungs using oxygen and manual compression to provide a tidal volume (Vt) exceeding baseline Vt, and using a Vt that is 50% greater than that delivered by the ventilator, requiring a peak inspiratory pressure of from 20 to 40 cm H2O.13 Four factors are considered important in performing the MH technique: the application of larger-than-normal Vt breaths,14,15 use of a slow inspiratory flow rate,16,17 an inspiratory pause,18,19 and a pressure manometer.20,21 In addition, the quick release of pressure on expiration leading to a rapid flow of air can simulate the effect of a cough.11,22,23 Even though no comprehensive studies have been done that incorporate and evaluate all four of the important MH techniques, there is support from clinical and research literature on the theoretical foundations and effectiveness of each of the factors separately. First, the use of larger-than-normal Vt is based on the hypothesis that by delivering a largervolume breath over time, MH may increase the expiratory flow rate and assist in moving secretions toward more proximal airways, where they can be cleared by suctioning.24 Second, the rate of inflation of the lung as a whole is a function of inflation pressure, compliance, and airway resistance. Nunn25 described the response to passive inflation of the lungs by the development of a constant airway pressure. If a constant inflation pressure is maintained, an alveolus with half the compliance but twice the resistance of another alveolus will increase in volume by half the volume change of the other alveolus. Thus, the relative distribution of gas between the two alveoli is independent of the rate or duration of inflation. In addition, using both hands to compress the bag can produce a Vt that is 50% greater than that delivered by the ventilator.26 Furthermore, the rate at which the bag is compressed, rather than the resistance of the circuit itself, is the main influence on the peak inspiratory flow rate.16,17,27 A fast inflation rate that does not allow the reservoir bag to fill adequately, and reduces the fraction of the inspired oxygen (Fio2).28,29 Moreover, delivering an increased Vt via MH may generate adequate transpulmonary pressure gradients to overcome alveolar atelectasis. Atelectatic alveoli do not reexpand immediately when the ventilator cycles with the inspiratory phase because a variable period of time is required before the alveolar critical opening pressure is reached.30 Therefore, the third important factor, the use of an inspiratory hold during www.chestjournal.org
MH, is thought to maintain these pressure gradients for an appropriate length of time. This technique may influence the distribution of the ventilation25 and allow time for alveolar inflation or enlargement, as well as the recruitment or unfolding of interdependent atelectatic alveoli. Finally, a pressure manometer can improve the performance of MH and optimize both the safety and the effectiveness of the treatment.31 Although there is no consensus about specific safe upper limits for peak airway pressure, barotraumas manifest at peak airway pressures of 26 to 64 cm H2O as demonstrated by several animal studies.20 Therefore, it is reasonable and prudent to minimize the peak airway pressure as much as possible during MH or any other ventilatory support procedure.32 This study examines the effect of MH in patients with atelectasis associated with ventilation support. The foundation for the practice of MH in this study is based on the best evidence available from clinical and research literature, and incorporates all four factors of MH technique considered to be important, as described above. The hypothesis of this study is that those in the experimental group should have improved sputum contents (wet/dry weight ratio, viscosity), respiratory system capacity (spontaneous Vt), maximal inspiratory pressure (Pimax), an improved rapid shallow breathing index (f/Vt), as well as improved chest radiograph signs and oxygenation ratio (Pao2/Fio2).
Materials and Methods Patient Selection Thirty-three patients with atelectasis due to ventilatory support were recruited at the pulmonary ICUs from two hospitals; of these 33 patients, 23 completed all of the study procedures. This study was a two-group, prospective, randomized study lasting 6 days. Patients were assigned to one of two groups: standard care with supplemental MH (experimental group, n ⫽ 10) or standard care only (control group, n ⫽ 13). The Institutional Ethical Committee approved the study protocol, and all patients gave informed written consent. The inclusion criteria consisted of the following: age ⱖ 40 years; ventilation support ⬎ 7 days and a positive end-expiratory pressure from 6 to 8 cm H2O; pulmonary atelectasis; excessive secretions (⬎ 30 mL/d); and spontaneous Vt ⬍ 250 mL and/or Pimax ⬍ 25 cm H2O and/or Vt ⬍ 400 mL under ventilator assistance. Pulmonary atelectasis was diagnosed using the following: (1) chest radiography showing increased infiltration, and (2) physical examination revealing weakness or muteness of the sounds in the involved area. The structural changes that develop in atelectasis increase the density of the lungs. The increase in lung density resists radiograph penetration and is revealed on radiograph films as increased opacity (ie, whiter in appearance). Thus, the more severe the atelectasis, the denser the lungs, and the whiter the radiograph film. Patients were assessed clinically and with a chest radiograph at recruitment to ensure the absence of a Fio2 ⱖ 0.6 requirement, CHEST / 128 / 4 / OCTOBER, 2005
2715
pulmonary pathology (for example, ARDS), active infection, acute cardiovascular dysfunction, or other systemic diseases. The trial took place between January 2001 and June 2001. There were three different types of ventilation systems used throughout this study (models 7200, 740, or 760; Nellcor Puritan Bennett; Temecula, CA). Standard Care All subjects were asked to continue any current prescribed medication (such as anticholinergic inhaled agents, inhaled corticosteroids, theophylline, prednisolone, or erythromycin) and chest physiotherapy (such as chest percussion, positioning, and suction) throughout the experiment. For the control group, these were the only prescribed treatments. None of the subjects received any sedation or narcotics. MH To ensure that uniform and correct techniques were employed, MH was administered by only one investigator. A 2.0-L reusable manual resuscitator (model 2153 MR100 plus; Galemed Corporation; Taipei, Taiwan) was used to deliver the MH breaths, and was connected to a flow of 100% oxygen at 15 L/min (calibrated with an oxygen analyzer). A force meter (Inspiratory Force Meter; Boehringer Laboratories; Norristown, PA) was connected between the resuscitator and the patient. Patients received MH to a peak airway pressure of 20 cm H2O by use of the resuscitator. The resuscitator was slowly compressed with both hands, and an inspiratory breath was maintained for 3 to 5 s at the end of pressing half of the resuscitator, and then completely pressing the resuscitator. Expiration was passive and unobstructed to facilitate expiratory flow with no positive end-expiratory pressure applied. Sufficient time was allowed for the resuscitator to fill completely prior to the next breath. Airway suctioning of the endotracheal tube was performed using size 14 catheters (Pahsco; Pacific Hospital Supply; Taipei, Taiwan) at the end of the MH procedure. The MH procedure was carried out at a rate of 8 to 13 breaths/min for a period of 20 min for each session tid (at 7:40 am, 11:40 am, and 3:40 pm) for 5 days on days 1 to 5 of the study. Sputum Sampling Nurses were instructed to collect and record the total amount of daily sputum (milliliters per 24 h) throughout the study. An aliquot of sputum from each patient’s total amount of daily sputum was freeze-dried (at – 80°C, at a negative pressure of 40 cm H2O) overnight to measure the wet/dry weight ratio. The viscosity of the sputum was measured using a viscometer at room temperature (25°C) with distilled water as a control, using sputum sampled by nurses at 7 am on days 0 (baseline), 3, and 6 of the study. After receiving chest percussion, the sputum was collected by airway suctioning of the endotracheal tube into a sterile pot. Measurement of Respiratory System Capacity Respiratory system capacity measurements were obtained 30 min after sputum sampling and just prior to the 7:40 am MH on days 0 (baseline), 3, and 6 of the study. The spontaneous Vt score was measured during ventilation disconnection (Haloscale Wright Respirometer; Ferraris Medical Limited; Middlesex, UK) by a respiratory therapist as spontaneous respiratory volume (milliliters) per minute divided by respiratory rate per minute. The Pimax was measured during ventilation disconnection with the inspiratory force meter by the respiratory therapist. A 2716
unidirectional expiratory valve pressure-manometer was connected to the endotracheal tube or tracheostomy, the port was occluded at end-expiration for 20 s, and after three spontaneous maximal inspiratory efforts the Pimax was recorded.33,34 The f/Vt score was measured during mechanical ventilation as a calculation of the ratio of the respiratory rate per minute (frequency) to the Vt setting (liters) from the display on the ventilation. A portable radiograph machine was used, and a staff radiologist reported the chest radiograph findings each morning. Scores were given as 1 (improved) or 2 (not improved). Nurses also recorded the ventilator volume and measured the cuff pressure (Control-inflator; VBM Medizintechnik; Sulz am Neckar, Germany) at least once or twice daily for patients with tracheostomy. The volume needed to attain a full seal should be recorded at least once or twice daily. The need for increasingly larger volumes indicates an expanding trachea. The pressure was kept at levels ⬍ 20 mm Hg. If there was an air leak in the cuff or cuff inflation system, nurses reinflated the cuff via a stopcock. If the ventilator could be set to compensate for the leak, the patient was not reintubated. If significant aspiration or inadequate ventilation was present, a new tube was inserted. Oxygenation Ratio Pao2/Fio2 was measured during mechanical ventilation as derived from arterial blood gas analysis and the Fio2 on days 0 (baseline) and 6 of the study. A calibrated blood gas analyzer (model 278; Ciba-Corning; Medfield, MA) was used for arterial blood gas analysis, and the Fio2 was read from the display on the ventilator obtained just prior to MH. Statistical Analysis Statistical software (version 10.0; SPSS; Chicago, IL) was used for data analysis. 2 test and Fisher Exact Test were used to assess the success of the randomization process in achieving two comparable groups. A t test was performed to establish the baseline stability of the dependent variables. A repeated-measures analysis of variance (ANOVA) was performed to compare scores over time between the experimental and the control groups on each of the seven dependent variables measured at each of the three time points: day 0 (baseline), day 3, and day 6 of the study. This method accounted for six covariates: sex (male, female), age, setting (medical center, local hospital), intubation (endotracheotomy, endointubation), logarithm of length of ventilation prior to enrolment, and logarithm of total sputum amount.35 The null hypothesis is that there is no interaction between group and study duration, ie, the 5 days of treatment with repeated measures on days 0, 3, and 6. There is a gradual increase in treatment effect if there is interaction between the group and the duration of the treatment. The Mantel-Haenszel 2 test for categorical data, adjusted odds ratio (OR), and multiple logistic regression were performed on the chest radiograph scores. The OR was calculated as the odds in favor of clinical improvement in the treatment group divided by the odds in favor of clinical improvement in the control group. Significance was indicated at p ⬍ 0.05.
Results Sample Initially, 33 patients agreed to participate in this investigation; of these, 10 patients withdrew. Twenty-three patients (n ⫽ 23) completed the full course Clinical Investigations in Critical Care
of treatment. The attrition rate in this experiment was high (30%), perhaps because of the characteristics of respiratory failure from mechanical ventilator support. Of the 10 patients who dropped out, 3 died (1 from the experimental group and 2 from the control group), 4 withdrew voluntarily (3 from the experimental group and 1 from the control group), and 3 received oxygen content ⬎ 50% during the study period (1 from the experimental group and 2 from the control group). As most of the withdrawals were for medical reasons, there is reason to hypothesize that some of these patients might have benefited from MH. In addition, there were no statistically significant differences in baseline demographics, clinical characteristics, or outcome measurements between those who withdrew and the remaining participants. Patient Characteristics Table 1 lists the sample baseline demographic and clinical characteristics including intubation, setting, and length of ventilation prior to enrollment. There were no statistically significant differences between the subjects of the two groups. The sample contained more men (n ⫽ 17) than women (n ⫽ 6); 74% of the subjects were ⬎ 65 years of age, 100% were married, and 74% had no history of smoking. All subjects were receiving mechanical ventilation for at least 7 days prior to study entry. The average Fio2 was 35%; pneumonia was diagnosed in 17 persons, and lowerlobe atelectasis was found in 19 persons.
Table 1—Characteristics of Subjects* Characteristics
Experimental Control All p Group Group Subjects Value†
Total subjects, No. 10 Sex Male 8 (80) Female 2 (20) Age, yr ⱖ 65 7 (70) ⱕ 64 3 (30) Cigarette history No 6 (60) Yes 4 (40) Intubation Tracheostomy 3 (30) Endotracheal tube 7 (70) Setting Medical center 8 (80) Local hospital 2 (20) Length of ventilation prior to enrollment, d 7 5 (50) 8 to 13 2 (20) ⱖ 14 3 (30)
13
Outcome Measures Outcome measure scores are compared in Table 3, and the mean scores and the p value for their group ⫻ time interaction in repeated-measures ANOVAs adjusted for covariates are listed. The spontaneous Vt and chest radiograph scores show significant differences between the experimental and control groups, and the f/Vt and Pao2/Fio2 scores show a trend toward improvement in the experimental group compared to the control group. In the experimental group, spontaneous Vt scores of 196.3 mL at baseline increased to 270.5 mL on day 6 (indicating an improvement) compared to the control group, which increased from 208.49 mL at baseline to 220.14 mL on day 6 (p ⫽ 0.035; Fig 1). Furthermore, in the experimental group, f/Vt scores of 216.59 at baseline decreased to 150.21 on day 6 compared to the control group, which decreased from 174.04 to 164.74 (p ⫽ 0.066; Fig 2). Moreover, in the experimental group, Pao2/Fio2 scores increased from 222.07 at baseline to 264.45 on day 6 compared to the control group, which decreased from 228.64 to 203.53 (p ⫽ 0.061; Fig 3). Further, in the experimental group, chest radiograph scores improved 15.55-fold (95% confidence interval, 1.14 to 239.77; p ⫽ 0.040) after adjustment for covariates
23 0.46
9 (69) 4 (31)
17 (74) 6 (26) 0.54
10 (77) 3 (23)
0.20 11 (85) 2 (15)
17 (74) 6 (26)
5 (39) 8 (61)
8 (35) 15 (65)
0.51
0.31 8 (61) 5 (39)
16 (70) 7 (30) 0.40
3 (23) 3 (23) 7 (54)
Table 2—Comparison of Baseline Measurements Between Groups (n ⴝ 23)*
17 (74) 6 (26)
8 (35) 5 (22) 10 (43)
*Data are presented as No. (%) unless otherwise indicated. †2 and Fisher Exact Test. www.chestjournal.org
Table 2 shows the actual values of the outcome variables in all patients at baseline. Differences in baseline values between the two groups were not statistically significant except for the sputum amount. The logarithm of the total sputum amount was used as covariates in the repeated-measures ANOVA model; therefore, the data from all the subjects were included in the analyses.
Variables Sputum content Amount, mL/24 h Wet/dry weight ratio, % Viscosity, min Respiratory system capacity Spontaneous Vt, mL Pimax, cm H2O f/Vt Oxygenation ratio Pao2/Fio2
Experimental Group (n ⫽ 10)
Control p Group (n ⫽ 13) Value†
186.50 ⫾ 116.57 27.82 ⫾ 11.48
94.62 ⫾ 53.64 19.29 ⫾ 13.67
0.02 0.13
9.67 ⫾ 17.91
8.57 ⫾ 12.60
0.87
196.30 ⫾ 80.87 208.49 ⫾ 54.56 27.00 ⫾ 16.87 20.92 ⫾ 10.46 216.59 ⫾ 146.62 174.04 ⫾ 66.62
0.67 0.30 0.36
222.07 ⫾ 93.94
0.90
228.64 ⫾ 131.84
*Data are presented as mean ⫾ SD. †t test. CHEST / 128 / 4 / OCTOBER, 2005
2717
Table 3—Repeated-Measures ANOVA on Outcome Measures Between Experimental and Control Groups (n ⴝ 23)* Outcome Measures Sputum content Sputum wet/dry ratio, % Day 0 (baseline) Day 1 Day 2 Day 3 Day 4 Day 5 Sputum viscosity, min Day 0 (baseline) Day 3 Day 6 Respiratory system capacity Spontaneous Vt, mL Day 0 (baseline) Day 3 Day 6 Pimax, cm H2O Day 0 (baseline) Day 3 Day 6 f/Vt Day 0 (baseline) Day 3 Day 6 Chest radiographs Improved/not improved, No. Adjusted OR 95% confidence interval for OR Oxygenation ratio Pao2/Fio2 Day 0 (baseline) Day 6
Experimental Group (n ⫽ 10)
Control Group (n ⫽ 13)
27.82 ⫾ 11.48 28.36 ⫾ 11.32 32.01 ⫾ 19.08 25.83 ⫾ 9.25 48.76 ⫾ 51.59 37.22 ⫾ 35.99
19.29 ⫾ 13.67 19.38 ⫾ 12.65 20.52 ⫾ 13.95 20.27 ⫾ 18.00 19.89 ⫾ 16.07 18.87 ⫾ 14.28
9.67 ⫾ 17.91 5.85 ⫾ 4.98 5.49 ⫾ 5.35
8.57 ⫾ 12.60 13.01 ⫾ 21.08 24.60 ⫾ 23.95
196.30 ⫾ 80.87 287.07 ⫾ 120.03 270.50 ⫾ 98.65
208.49 ⫾ 54.56 223.61 ⫾ 63.26 220.14 ⫾ 79.34
27.00 ⫾ 16.87 30.30 ⫾ 10.26 36.10 ⫾ 16.16
20.92 ⫾ 10.46 17.92 ⫾ 9.54 18.38 ⫾ 8.14
216.59 ⫾ 146.62 133.67 ⫾ 84.08 150.21 ⫾ 66.12
174.04 ⫾ 66.62 144.22 ⫾ 65.55 164.74 ⫾ 101.50
9/1 16.56 1.14–239.77
6/7 1.00
222.07 ⫾ 93.94 264.45 ⫾ 113.41
228.64 ⫾ 131.84 203.53 ⫾ 96.17
p Value† 0.831
0.145
0.035
0.194
0.066
0.040
0.061
*Data are presented as mean ⫾ SD unless otherwise indicated. †Treatment ⫻ time interaction in repeated-measures ANOVA, adjusted for sex, age, setting, intubation, logarithm of length of ventilation prior to enrollment, and logarithm of total sputum amount.
Figure 1. Mean values of spontaneous Vt in both groups; horizontal bars ⫽ ⫹ 1 SD. p ⫽ 0.035 refers to differences between groups over time, with changes only on day 3 and day 6 while adjusting for sex, age, setting, intubation, logarithm of length of ventilation prior to enrollment, and logarithm of total sputum amount. Higher spontaneous Vt values represent patients with atelectasis associated with ventilation support-improved alveolar recruitment. 2718
Figure 2. Mean values of f/Vt in both groups; horizontal bars ⫽ ⫹ 1 SD. p ⫽ 0.066 refers to differences between groups over time, with changes only on day 3 and day 6 while adjusting for sex, age, setting, intubation, logarithm of length of ventilation prior to enrollment, and logarithm of total sputum amount. Lower f/Vt values represent patients with atelectasis associated with ventilation support-improved alveolar recruitment. Clinical Investigations in Critical Care
Figure 3. Mean values of Pao2/Fio2 in both groups; horizontal bars ⫽ ⫹ 1 SD. p ⫽ 0.061 refers to differences between groups over time, with changes only on day 6 while adjusting for sex, age, setting, intubation, logarithm of length of ventilation prior to enrolment, and logarithm of total sputum amount. Higher Pao2/ Fio2 values represent patients with atelectasis associated with ventilation support-improved alveolar recruitment.
compared with the control group. Scores in sputum wet/dry weight ratio increased from 27.82% at baseline to 37.42% on day 6 in the experimental group and decreased from 19.29 to 18.87% in the control group; scores in sputum viscosity decreased from 9.67 min at baseline to 5.49 min on day 6 in the experimental group, and increased from 8.57 to 24.60 min in the control group, Pimax scores of 27.00 cm H2O at baseline increased to 36.10 cm H2O on day 6 in the experimental group, and decreased from 20.92 to 18.38 cm H2O in the control group after adjustment for covariates compared with the control group, even though these changes were not statistically significant.
Discussion To our knowledge, this is the first study to examine the potential benefits of MH to 20 cm H2O by oxygen-powered, manual resuscitation bag with an inspiratory breath-hold of 3 to 5 s, while pressing half of the resuscitator, in a group of intubated patients with atelectasis. In addition, this study employed methodologic features that strengthened validity and reliability of the findings, including the randomization of subjects to groups and multivariate analysis controlling for known covariates. Those receiving MH had statistically significant www.chestjournal.org
improvement in respiratory system capacity and oxygenation ratio, evidenced by the scores of spontaneous Vt and chest radiograph signs, and a trend toward improvement of f/Vt and Pao2/Fio2 compared to the control groups, after adjusting for the effects of sex, age, setting, intubation, logarithm of length of ventilation prior to enrolment, and logarithm of total sputum amount. These results are not consistent with the fact that the MH technique was initially designed to enhance clearance of airway secretions.27 Nevertheless, it supported the initial hypothesis of this study, that MH improves alveolar recruitment by delivering a larger-volume breath over time,24 and by the development of a constant airway pressure25 in patients with atelectasis from mechanical ventilation. In addition, MH produced no adverse events in the experimental group, as none of the patients experienced pneumothorax, suffocation, or hypotension during or following MH. However, potential limitations of this investigation should be considered when interpreting the findings. These include the following: many of the outcome measures have a subjective component to them, because the respiratory therapist who scored these outcomes was not blinded; failure to obtain outcome data of successfully weaned patients before they completed this study; length of ventilation; and the small sample size. In addition, this model did not account for cigarette consumption, length of atelectasis, and the severity of the illness, all of which could account for group differences over time. Future studies should incorporate such risk adjustment using standard severity of illness measures as APACHE (acute physiology and chronic health evaluation)36 or sequential organ failure assessment37 scores. Further study on the effects of MH should be conducted using different resuscitation circuits (such as MH to 30 or 35 cm H2O), different subject groups (such as receiving ventilatory support for ⬍ 7 days, and differentiating between chronic and acute atelectasis), and different operators (such as physiotherapists, respiratory therapists, or nurses). Additional studies are also needed to elucidate the long-term outcomes such as time to extubation, time to discharge, ventilator-free days, and discharge status (home, long-term pulmonary care, death). Specifications by the practitioner, and patient preferences for treatment duration and frequency should also be explored.
Conclusion This study provides evidence that MH performed in a stable patient with atelectasis associated with CHEST / 128 / 4 / OCTOBER, 2005
2719
ventilation can improve spontaneous Vt and chest radiograph signs, and a trend toward improvement of f/Vt and Pao2/Fio2. MH is a nursing intervention that could be implemented without a physician’s order, and has the potential to make a positive impact on patient outcome. While this preliminary study with a small sample size does not warrant changes in clinical practice at this time, it does contribute to the evidence base on the benefits of MH in critically ill and ventilator-dependent patients. Further investigations are required to replicate this study with a larger sample size, evaluate different techniques of MH, evaluate effects of MH in different patient groups, and determine the longterm outcomes of MH. As we continue to build evidence through additional studies, we may eventually be able to recommend practice guidelines for the procedure of MH for health professionals to treat various clinical conditions. ACKNOWLEDGMENT: The authors like to express their appreciation to Ivo L. Abraham, PhD, RN, of Matrix45, LLC, and the School of Nursing, University of Pennsylvania, Philadelphia, PA, and Karen M. MacDonald, PhD, RN, of Matrix45, LLC, Earlysville, VA, for the time and effort they spent commenting on earlier versions of this article. The authors also like to acknowledge the contribution of the physiotherapy, nursing, and medical staff of Chang Gung Memorial Hospital for their expert assistance, especially Dr. Han-Pin Kuo.
References 1 Konrad F, Schreiber T, Brecht-Kraus D, et al. Mucociliary transport in ICU patients. Chest 1994; 105:237–241 2 Anderson J, Jenkins S. Physiotherapy problems and their management. In: Webber B, Pryor J, ed. Physiotherapy for respiratory and cardiac problems. London, UK: Churchill Livingstone, 1993; 226 –227 3 Berney S, Denehy L. A comparison of the effects of manual and ventilator hyperinflation on static lung compliance and sputum production in intubated and ventilated intensive care patients. Physiother Res Intern 2002; 7:100 –108 4 Wagner DP. Economics of prolonged mechanical ventilation. Am Rev Respir Dis 1989; 140(suppl):14S–18S 5 Hodgson C, Denehy L, Ntoumenopoulos G, et al. An investigation of the early effects of manual lung hyperinflation in critically ill patients. Anesth Intensive Care 2000; 28:255–261 6 McCarren B, Chow CM. Description of manual hyperinflation in intubated patients with atelectasis. Physiother Theory Pract 1998; 14:199 –210 7 Rothen HU, Sporre B, Engberg G, et al. Re-expansion of atelectasis during general anesthesia: a computed tomography study. Br J Anaesth 1993; 71:788 –795 8 Stiller K, Geake T, Taylor J, et al. Acute lobar atelectasis: a comparison of two chest physiotherapy regimens. Chest 1990; 98:1336 –1340 9 Barker M, Adams S. An evaluation of a single chest physiotherapy treatment on mechanically ventilated patients with acute lung injury. Physiother Res Intern 2002; 7:157–169 10 Patman S, Jenkins S, Stiller K. Manual hyperinflation: effects on respiratory parameters. Physiother Res Intern 2000; 5:157–171 2720
11 Jones AYM, Hutchinson RC, Oh TE. Effects of bagging and percussion on total static compliance of the respiratory system. Physiotherapy 1992; 78:661– 666 12 Tweed WA, Phua WT, Chong KY, et al. Tidal volume, lung hyperinflation and arterial oxygenation during general anaesthesia. Anaesth Intensive Care 1993; 21:806 – 810 13 Denehy L. The use of manual hyperinflation in airway clearance. Eur Respir J 1999; 14:958 –965 14 Maxwell L, Ellis E. Secretion clearance by manual hyperinflation: possible mechanisms. Physiother Theory Pract 1998; 14:189 –197 15 McCarren B, Chow CM. Manual hyperinflation: a description of the technique. Aust J Physiother 1996; 42:203–208 16 Baker AB, Colliss JE, Cowie RW. Effects of varying inspiratory flow waveform and time in intermittent positive pressure ventilation: various physiological variables. Br J Anaesth 1977; 49:1221–1233 17 Pillet O, Choukroun ML, Castaing Y. Effects of inspiratory flow rate alterations on gas exchange during mechanical ventilation in normal lungs: efficiency of end-inspiratory pause. Chest 1993; 103:1161–1165 18 Bindslev L, Santesson J, Hedenstierna G. Distribution of inspired gas to each lung in anesthetized human subjects. Acta Anaesth Scand 1981; 25:297–302 19 Dammann J, McAslan T, Maffer C. Optimal flow pattern for mechanical ventilation of the lungs. Crit Care Med 1978; 6:293–310 20 Redfern J, Ellis E, Holmes W. The use of a pressure manometer enhances student physiotherapists’ performance during manual hyperinflation. Aust J Physiother 2001; 47: 121–131 21 Hila J, Ellis E, Holmes W. Feedback withdrawal and changing compliance during manual hyperinflation. Physiother Res Intern 2002; 7:53– 64 22 Hack I, Katz C, Eales C. Airway pressure changes during “bag squeezing.” S Afr J Philos 1980; 36:97–99 23 Windsor H, Harrison G, Nicholson T. “Bag squeezing”: a physiotherapeutic technique. Med J Aust 1972; 2:829 – 832 24 Clement AJ, Hubsch SK. Chest physiotherapy by the ‘bag squeezing’ method. Physiotherapy 1968; 54:355–359 25 Nunn JF. Nunn’s applied respiratory physiology. 5th ed. Oxford, UK: Butterworth-Heinemann, 2000; 166 26 Chulay M, Graeber GM. Efficacy of a hyperinflation and hyperoxygenation suctioning intervention. Heart Lung 1988; 17:15–22 27 Maxwell L. Ellis ER. The effects of three manual hyperinflation techniques on pattern of ventilation in a test lung model. Anaesth Intensive Care 2002; 30:283–288 28 Eaton JM. Adult manual resuscitators. Br J Anaesth 1984; 31:67–70 29 Glass C, Grap M, Corley M, et al. Nurses’ ability to achieve hyperinflation and hyperoxygenation with a manual resuscitation bag during endotracheal suctioning. Heart Lung 1993; 22:158 –165 30 Galvis AG, Bowen A, Oh KS. Nonexpandable lung after drainage of pneumothorax. Am J Respir Crit Care Med 1981; 136:1224 –1226 31 Goldstein B, Catlin EA, Vetere JM, et al. The role of in-line manometers in minimizing peak and mean airway pressure during hand regulated ventilation of newborn infants. Respir Care 1989; 34:23–27 32 Haake R, Schlichtig R, Ulstad D, et al. Barotrauma: pathophysiology, risk factors, and prevention. Chest 1987; 91:608 – 613 33 Marini JJ, Smith TC, Lamb VJ. Estimation of inspiratory muscle strength in mechanically ventilated patients: the Clinical Investigations in Critical Care
measurement of maximal inspiratory pressure. J Crit Care 1986; 1:32–38 34 Caruso P, Friedrich C, Denari SDC, et al. The unidirectional valve is the best method to determine maximal inspiratory pressure during weaning. Chest 1999; 115:1096 –1101 35 Tukey JW. Exploratory data analysis. Reading, MA: AddisonWesley, 1977
www.chestjournal.org
36 Knaus WA, Draper EA, Wagner DP, et al. APACHE II: a severity of disease classification system. Crit Care Med 1985; 13:818 – 829 37 Vincent JL, Moreno R, Takala J, et al. The SOFA score to describe organ dysfunction/failure. Intensive Care Med 1996; 22:707–710
CHEST / 128 / 4 / OCTOBER, 2005
2721