Respiratory Responses to Chest Compression in Human Subjects1 , 2

TAKASHI NISHINO, TERUHIKO ISHIKAWA, ATSUKO TANAKA, and KAZUAKI HIRAGA

Introduction

Studies of responses to experimental loads added to the airway of animals and humans suggest that the respiratory system has a marked ability to maintain an adequate alveolar ventilation while preserving oxygen and carbon dioxide homeostasis. The essence of the respiratory control system is feedback of information concerning the effectiveness of the respiratory response. In the face of respiratory loading, the central and peripheral chemoreceptors must play a dominant role in that gaseous homeostasis is preserved and proprioceptive information from the impeded respiratory muscle takes the role of determining how it is accomplished. The response of anesthetized humans to experimental loads has been investigated in several studies (1-6). However, there are some conflicting results showing complete compensation (1, 2) and incomplete compensation (3-5) during threshold and flow-resistive loads. The recent study of Moote and colleagues (6) showed that steady-state compensation for both flow-resistive and elastic inspiratory loads was related to the magnitude of the initial loads. Thus, for loads causing about a 30070 reduction in the first loaded tidal volume, steady-state compensation was virtually complete, whereas for loads causing a 50% or greater reduction in the first loaded tidal volume, compensation was incomplete. External ventilatory loading produced by chest compression has been used by several investigators to study the involvement of lung and chest wall reflexes, which play an important role in respiratory load compensation (7-10). These studies have shown that restriction of chest cage expansion or reduction in the volume ofthe chest in awake humans increases the frequency of breathing and maintains adequate ventilation, indicating the presence of complete compensation. Most of the previous studies, however,except for the study of Agostoni and coworkers (9), were concerned with the effects of chest compression on breath980

SUMMARY Totest the hypothesis that respiratory compensation for chest compression In the conscious state Is different from that In the unconscious state, a range of chest compression was applied to eight conscious and eight anesthetized subJects. Chest compression was carried out by Inflating a pneumatic cuff Inserted under a chest corset and placed over the anterior chest wall for 5 min. Cuff Inflation pressures ware 25, 50, and 100 mm Hg: light, moderate, and heavy chest compression, respectively. In both conscious and anesthetized subjects, chest compression Immediately caused a decrease In tidal volume (VT) and a concomitant Increase In respiratory frequency (f). These changes stabilized within 2 min and remained nearly steady for the rest of the sustained chest compression. The decrease In VT was more prominent the greater the loads, and the change In f was In reverse proportion to the change In VT. In conscious subjects the Increase In f was due to shortening of both Inspiratory time (TI) and expiratory time (TE),whereas In anesthetized subjects the Increase In f was due solely to shortening of TE. In conscious subjects, the values of minute ventilation (VI) and end-tidal peo, (PETeo,) during light and moderate chest compression ware not different from the control values, whereas VI Increased and PETeo,decreased during heavy chest compression. In anesthetized subjects, the values of VI and PETeo, were not different from the control values at any different magnitudes of chest compression. Our results Indicate that both conscious and aneethetlzed subjects respond reflexively to chest compression In a load-related manner but that there Is a slight difference In respiratory response to chest compression betwaen the conscious and the anesthetized state. AM REV RESPIR DIS 1992; 148:980-984

ing patterns in the conscious state, and the respiratory responses to chest compression during anesthesia have not been fully studied. The purpose of the present study is to characterize the responses to a range of sustained chest compression in anesthetized humans and to compare these responses with the responses in conscious subjects. Methods The studies wereperformed in the supine position on eight volunteers (consciousness group) aged 29.3 ± 2.0 yr and eight female patients scheduled for elective hysterectomy (anesthesia group) aged 39.8 ± 2.1 yr (mean ± standard error of the mean, SEM). The average heights and weights of the consciousness group were 159.3 ± 1.2 em and 49.5 ± 1.2kg, and those of the anesthesia group were 154.4 ± 1.6 em and 50.9 ± 2.7 kg (mean ± SEM), respectively. None had clinicalevidence of respiratory, cardiovascular, or neuromuscular disorders. The protocol was approved by the ethics committee of our institution, and each subject gave informed consent. In the consciousness group, the subject was fitted with a face mask and breathed 40070 O 2 in N2 from a non breathing circuit incorporating a Fleisch pneumotachograph (No. 2; Dynasciences, Blue Bell, PA). Ventilatory airflow (Y) was measured with the pneu-

motachograph and a differential pressure transducer (TP-602T, Nihon Koden); and tidal volume (VT) wasobtained byelectricalintegration of inspired flow. Airway pressure (Paw) was measured with a pressure transducer (DDL-0.05; Toyo Baldwin). End-tidal Pco, (PETco.) was.monitored continuously with an infrared CO, analyzer (Norrnocap'[; Datex), Heart rate (HR) and arterial oxygensaturation were continuously measured with an HR counter (AT-61OG; Nihon Koden) connected to an electrocardiographic (BCG)monitor and with a pulse oximeter (BioxilP 3740; Ohmeda), respectively. Y, VT,Paw, PETeo" ECG, and HR all were recorded on a Nihon Koden eight-channel recorder (RJG-4128). A cloth corset (25 ern wide) with adjustable straps was placed around the chest, and a rectangular pneumatic cuff (18 x 36 ern) was inserted under the corset and placed over the anterior chest. The straps of the corset

(Received in original form December 3, 1991 and in revised form March 10. 1992) 1 From the Department of Anesthesiology, National Cancer Center Hospital, Tokyo, Japan. 2 Correspondence and requests for reprints should be addressed to Dr. T. Nishino, Department of Anesthesiology, National Cancer Center Hospital. Tsukiji 5-1-1, Chuo-ku, Tokyo 104, Japan.

981

RESPIRATORY COMPENSATION FOR CHEST COMPRESSION

(A)

3 min

,--,

HR ECG

H+I+++++H+++++++-IIIIII'IIII" 1111111111111111111111111'111111,

~Off

1111111 1I11111111111111111111111111111111111-H++H++++++++mIH+I+h

em H20

Paw

19c:

AF

19[ ~..-/"'-~"---'~~~~~~~~~

VT

°i IJJ j.JJJJJJJLfL!ULJLJULJlJL

lis

jJJJJJJJJJJJjJJJJJJ:.-

mmHg

PET CO,

SO[ 0

(8) HR BP

'JijJUlijilUUlj!ltiflJUUIil!HijtllUUlUi'mUJJjtl!ljjilWJjjij.lUij.I,J~W~

em H20

19[ ~-o,~-~~~~---~-

Paw

AF I

VT

02~[ IJ~JJJJJJJJJJJJJJJ PETCO,

JJJJJJJJ~JJJJJJjJJJ~

~rl[1J'JV'vVVV\flJlJlJl[Vl\ rv\flrljljlII[V~VlflJl!VVVlr\[l

Fig. 1. Experimental records illustrating respiratory responses to chest compression (Pcuff = 100 mm Hg) in a conscious subject (A) and in an anesthetized subject (8). Arrows indicate the start and end of chest compression. HR, heart rate; ECG, electrocardiogram; Bp, arterial blood pressure; Paw, airway pressure; AF, airflow; vi, tidal volume; PETeo" end-tidal peo,.

wereadjusted so that breathing was not hampered when the pneumatic cuff was deflated. When all respiratory variables werestable, the pneumatic cuff was inflated in random sequence using an Inflatolvlatic" Regulator 3000 (Zimmer, Indiana) to a pressure of 25, 50, and 100mm Hg. Each chest compression was maintained for 5 min and was carried out in random sequence with an interval of 5 min. The measurements of respiratory variables weredone beforethe start of chestcompression (control period), during chest compression, and after the releaseof chest compression (recoveryperiod). Also, after completion of these procedures, the face mask of each conscious subject was connected to a 9-L water-sealed spirometer (Aika) and the decreases in lung volume from FRC in response to each chest compression was measured. In the anesthesia group, all the subjects were intramuscularly premedicated 1 h before the induction of anesthesia with 0.5 mg atropine and 50 mg of hydroxyzine. Anesthesia was induced with intravenous tlunitrazepam (0.04 mg/kg) and intravenous pentazocine (0.3 mg/kg) followed by intravenous succinylcholinechloride (1 mg/kg). A cuffed endotracheal tube (7.5 mm inner diameter) was then inserted after spraying the larynx and the trachea with 4070 lidocaine. Subsequently, anesthesia was maintained with 60070 NzO in Oz, and spontaneous respiration was allowed to resume. After spontaneous respiration returned, the endotracheal tube of each patient was connected to the experimental apparatus. In addition to the monitoring of HR and arterial oxygen saturation, arterial blood

pressure (BP) via a radial artery catheter was continuously monitored with a pressure transducer (P23 ID; Gould-Statham). The response to chest compression was measured in the same manner as performed in conscious subjects. Also, during the control period and in the period immediately before the release of chest compression, the inspiratory line was occluded at end expiration for one breath and the peak of negative pressure (occlusion pressure Pmax) was obtained. Values of respiratory and circulatory variables during the control, at the end of chest compression, and immediately after removal of chest compression were obtained from measurements of 30-s periods. Inspiratory time (11) is defined as the time between the onset of inspiratory flow and the onset of expiratory flow, and expiratory time (Th)is defined as the remainder of the cycle. Minute ventilation (VI) is defined as the product of VT and respiratory frequency (f). Statistical analysis was performed using two-way analysis of variance, followed by Dunnett's test. A p < 0.05 was considered significant.

Results

Responses in Conscious Subjects In conscious subjects, chest compression immediately caused an increase in f with a decrease in VT (figure lA). These changes were proportional to the magnitude of chest compression and stabilized within 2 min after the start of chest compression. Thereafter the breathing

pattern remained nearly steady. In addition to changes in breathing pattern, HR usually increased over 2 min to a level that remained nearly steady for the remainder of chest compression. Also, immediately after the start of chest compression, a transient increase in PETC02 was occasionally observed, which gradually returned to the control level during sustained chest compression. Immediately after the removal of the chest compression, there was an overshoot of VT. The mean values of respiratory variables for the control period (C), steadystate period during chest compression (8), and the recoveryperiod (R) are shown in figure 2. Changes in VT and f were more pronounced the greater the degree of chest compression. The increase in f was due to shortening of both Tr and Th and was in reverse proportion to the decrease in VT (figure 3). Despite great changes in breathing pattern, the values of VI and PETC02 during light and moderate chest compression were not different from control values, whereas during heavy chest compression, VI increased and PETC02 decreased. Also, there was a significant increase in HR during heavy chest compression (table 1). The decreases in lung volume in response to cuff pressures of 25, 50, and 100 mm Hg were 128 ± 13, 275 ± 16, and 480 ± 25 ml (mean ± SEM), respectively. There were significant differences (p < 0.01) between these values.

Responses in Anesthetized Subjects Responses to chest compression in anesthetized subjects are essentially similar to those in conscious subjects. Like the responses in conscious subjects, chest compression immediately caused an increase in f and a decrease in VT, which stabilized within 2 min after the start of chest compression (figure lB). The changes in breathing pattern were more pronounced the greater the loads, and there was an inverserelationship between the increase in f and the decrease in VT (figure 3). However, unlike the responses in conscious subjects, the increase in f was due solely to shortening of Ts, and the values of VI and PETc02 were not different from control values with all degrees of chest compression (figure 4). Also, occlusion pressure during chest compression increased in proportion to the magnitude of chest compression. Changes in HR and BP are listed in table 1. HR increased in proportion to the magnitude of chest compression,

982

NISHINO, ISHIKAWA, TANAKA, AND HIRAGA

Ilmin

_pm

m'

10

20

700

VT

,+\ /f· "

re 800

ffiW '\ "... ·'····,/! j.")

10

s

TI

"

c

s

.

R

S

R

m..",

S

~,····,,·t········Att . "'..~~f .·:::