INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Infant pulmonary function testing has become a major focus of investigation as pediatric pulmonologists study the maturation of lung function in the infant and explore the impact of infant lung disorders on function of the developing lung (1). These tests are now being used to monitor the recovery of lung function in infants after neonatal or infant lung disease (4). Infant lung function testing is also employed to determine the effectiveness of various medications in the treatment of lung disorders, medications such as bronchodilators or diuretics, on lung compliance, resistance, and maximal flow characteristics of the lung (7). The method for measurement of expiratory flow limitation using thoracoabdominal compression (TAC) was originally described by Adler and Wohl (10) and frequently modified (12). TAC produces forced expiratory flow in sedated infants by sudden inflation of a vest surrounding the infant's chest and abdomen at the end of a tidal inspiration. With incremental increase in the vest pressure during subsequent chest and abdominal compressions, a pressure is usually defined at which maximal airflow is limited. The present study was conduced to determine whether TAC leads to alteration in lung mechanics. We hypothesize that TAC in infants and young children results in changes in chest wall mechanics and/or a reduction in lung volume, which will be reflected in a fall in total thoracic compliance.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

We studied forty-one infants born of HIV-infected mothers enrolled at five centers participating in a study of the longitudinal effects of vertically transmitted HIV on pulmonary and cardiac complications in infants and children (P²C² HIV Study) (19). This substudy was conducted as part of the larger study of the longitudinal changes in infant pulmonary function in all of the infants enrolled in the P²C² HIV Study. The study was approved by the institutional review boards (IRB) at each of the participating institutions. Informed consent was obtained from the parent or legal guardian of all study participants. Each infant was studied 4 to 6 h after the last feeding. Chloral hydrate in a dose of 80 to 100 mg/kg body weight (with a maximal dose of 1,000 mg of chloral hydrate) was used to induce sleep. Infants were studied supine with a roll placed under the neck and shoulders to maintain the head in "sniffing" position during the study. Oxygenation was monitored continuously during the study using a pulse oximeter with digital read-out of heart rate and oxygen saturation (Nellcor, Hayward, CA). Infant lung function was measured using the Sensormedics 2600 infant lung function cart (SensorMedics Corp., Yorba Linda, CA) and an appropriately sized face mask with a soft air-filled collar around its perimeter providing a leak-proof seal (Vital Signs, Totowa, NJ) over the nose and mouth. The face mask was connected to a pneumotachograph (10, 30, or 160 LPM) with a volume range of 0 to 127, 0 to 255 or 0 to 2048 ml (Hans Rudolph Inc., Kansas City, MO) depending on the infant's size and predicted maximal expired airflow. The pneumotachograph assembly included an occlusion-valve device (dead space, 5.0 ml). Passive mechanics (Crs, Rrs, Trs) were obtained as previously described by Mortola and colleagues (14) and LeSouef (15, 16). After a period of quiet breathing at end of a tidal inspiration the airway was occluded by the shutter for 200 ms to obtain a pressure plateau of at least 100 ms. A passive exhalation followed during which airway pressure and flow (integrated to volume) were recorded. The exhalation was considered passive if the resultant curve was straight and noise-free within 65 to 90% of exhaled volume. Crs was determined by computer, placing the points on the expiratory flow limb at 65 and 90% of tidal volume. Along with total thoracic compliance (Crs), the total pulmonary resistance (Rrs), and time constant of the respiratory system (Trs) were also calculated.

After the initial measurements of Crs, Rrs, and Trs, an inflatable vest was fastened around the chest and upper abdomen with the arms outside of the vest. Forced expiratory flow (17) was obtained by sudden, rapid chest and abdominal compression initiated at end-tidal inspiration to generate a partial maximal expiratory flow-volume curve. Rise time of the system, excluding the vest, was 50 ms. The initial compression pressure was 40 cm H₂O. A second trial was recorded at the same pressure. Then, the jacket pressure was increased in 10 cm H₂O increments. Two trials of TAC were obtained at each successive 10 cm H₂O pressure increment until flow limitation was achieved or a chest compression pressure of 100 cm H₂O was applied. After the TAC study (mean 10 [range, 6 to 14] compressions), Crs, Rrs, and Trs were measured again after opening the vest. The number of technically acceptable curves ranged from 6 to 20 (mean, 11.1) for the pre-TAC measurement, 4 to 11 (mean, 9.7) for the post-TAC measurement. Crs, Rrs, and Trs were calculated and reported in a similar manner to the pre-TAC determinations of these parameters.

Statistical Methods

Descriptive statistics were provided as the mean and standard deviation. Percentage change from baseline in Crs, Rrs, and Trs after TAC was compared using the one-sample t-test. Percentage change between groups (sex and HIV status) of study subjects was compared using a two-sample t-test and a one-way analysis of variance with a linear trend test for age. Confidence intervals (95%) were calculated for percentage change in Crs, Rrs, and Trs. Spearman's rank correlation coefficient was used to test for association between percentage change and age and percentage change and length of time between TAC and the pulmonary function measurement. Multiple linear regression analysis was used to examine the relationship between change in Crs, Rrs, and Trs and age, sex, clinical center, and HIV status.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Complete study data were obtained on 41 infants, 22 (54%) male and 19 (46%) female. Thirty-four percent of the study subjects were HIV-infected. The mean age of the infants studied was 12.4 ± 7.5 mo (range, 5.3 to 31.4 mo). Thirty-eight of the 41 infants were younger than 2 yr of age at the time of study. The means of pre and post-TAC coefficient of variation were 17.5 and 12.2 for Crs, 16.6 and 14.9 for Rrs, and 20.4 and 16.3 for time constant (Trs). The pre- with the post-TAC values for Crs, Rrs, and Trs are compared in Table 1. Both Crs and Trs means were significantly reduced after TAC (p = 0.013 and 0.003, respectively), whereas Rrs did not change significantly after the TAC study. However, changes in Crs and Trs were relatively small. The average post-pre TAC percent change in Crs was 8.2% (95% CI: 14.6 to 1.8%) and the average percent change in Trs was 11.6% (95% CI: 19.2 to 4.0%). The average change in Rrs was only 1.5% (95% CI: 8.6 to +5.7%). The mean time from TAC to postcompression measurement of Crs, Rrs, and Trs (available for 38 of the infants) was 11.0 ± 10.4 min (range, 2 to 30 min). The length of time between TAC and measurement of Crs, Rrs, and Trs was not associated with change in Crs, Rrs, or Trs (p > 0.7, Spearman's correlation coefficients). Age of the infant correlated significantly inversely with change in Crs (r = 0.39, p = 0.01) and Trs (r = 0.43, p = 0.005), but not with change in Rrs (r = 0.09).

Changes in Crs, Rrs, and Trs were also examined in relationship to three age categories (< 6 mo, 6 to 12 mo,  1 yr), sex, and the HIV status of the child (Table 2). Sex of the child was not significantly related to the change in Crs, Rrs, or Trs. We observed a significant trend in which larger mean percentage declines in Crs and Trs, but not Rrs, occurred with increasing age of the child. Similar results were obtained with Spearman's correlation analysis. HIV-infected children exhibited a significantly greater decline in Crs (p = 0.003) than did HIV-uninfected children, for whom the decline in Crs was not statistically significant (p = 0.64). The mean percentage change in Rrs was not significantly different from zero for either HIV-infected or HIV-uninfected children. However, HIV-infected children studied had an increase in mean Rrs after TAC, which differed from HIV-negative children in whom we observed a decline in Rrs (p = 0.03). Mean percentage change in Trs was negative, but not significantly different, for both HIV-infected and HIV-uninfected children. Plots of the postcompression versus precompression measurement of Crs, Rrs, and Trs by HIV status are shown in Figure 1.

View larger version (22K): [in this window] [in a new window]   Figure 1.   Plots of postcompression versus precompression measurement of Crs, Rrs, and Trs by HIV status.

Percent changes in Crs and Rrs were significantly related to HIV status of the child after adjusting for age and sex, whereas sex and age did not correlate with significant changes in Crs and Rrs. Trs was not significantly related to age, sex, or HIV status. Table a-cconfirmsthesefindingsandshowstheeffectof3aIV status, but not age, on change in Crs and Rrs, and the lack of an effect of age or HIV status on change in Trs. When clinical centers was included in these models, it was not a significant predictor of change in Crs or Trs, but it was significantly related to change in Rrs. However, adjustment for clinical center did not alter the basic findings with respect to HIV status, sex, or age for any of the models.

Because of the small number of infants with abnormal clinical respiratory findings (e.g., tachypnea, cyanosis, adventitious breath sounds) at the time of the study, we could not adequately assess whether there was a correlation between the finding of crackles or airway hyperreactivity on physical examination and a fall in Crs after TAC.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The thoracoabdominal compression technique to measure the limitation of expired flow in infants utilizes repeated compressions of the chest and upper abdomen with an inflatable vest. Increases in inflation pressure, in steps of 10 cm H₂O are performed to define the point at which there is no further increase in peak expiratory flow. TAC may entail as many as 14 to 16 sequential compressions of the infant's chest and abdomen until it is determined whether flow limitation has been achieved or, on the other hand, two trials at a TAC of 100 cm H₂O have been completed without achieving flow limitation. Infants, especially in the first year after birth, have a highly compliant chest wall because of lack of calcification in their sternum and ribs. Thus, the ribs and sternum provide little protection to the inflated lung against atelectasis after forceful deformation of the chest and abdominal wall. However, in this study we found that older subjects received, on the average, higher TAC pressures (Colin, et al., in preparation). TAC may still induce airway and chest wall deformation and/or changes in lung volume that might impact on other measurements of other lung function parameters. Ratjen and colleagues (17) found that Crs measured at volumes above tidal volume led to elevation of Crs suggesting recruitment of closed or atelectatic lung compartments. Studies by Patel and colleagues (18) revealed an improvement in Crs, but not Rrs, for as long as 20 min after the use of forced lung deflation (40 cm H₂O applied to the endotracheal tube) in intubated and ventilated infants with lung disease. The mean age of infants in the study of Patel and colleagues was 7 mo. They hypothesized that the change in Crs after application of negative pressure to the airway was as a result of reexpansion recruitment of atelectatic lung units resulting from assisted ventilation. In our study positive pressure was applied around the chest and abdomen in infants who were asleep and breathing spontaneously. Neither HIV-infected or uninfected subjects had evidence of respiratory insufficiency at the time of this study. We observed changes in respiratory system compliance that were age-dependent. The major changes in Crs occurred in those infants older than 6 mo of age, whereas change in resistance was not age-related, and the time constant for the respiratory system decreased only in those infants older than 1 yr of age. We have found, after adjustment for the impact of HIV infection, that there was not a significant age effect. Airway conductance and the time constant for the emptying of the lung (Trs) may be improved by augmentation of clearance of mucus from previously partially closed or obstructed airways.

On the other hand, in the older infant, application of pressure around the chest and abdomen may force air from stable, expanded portions of the lung into smaller, unexpanded units, which require a higher opening pressure. This would result in lowering of Rrs and improvement of Trs because of the enhanced retractive forces acting on the airways of expanded lung units.

We also found that HIV-infected infants participating in this substudy were for the most part without significant abnormal clinical respiratory findings at the time of study. However, they were more likely to experience decline in Crs (p = 0.001) than were uninfected infants (p = 0.64) after TAC. These findings suggest that even at an early age when they have no respiratory symptoms or clinical respiratory findings, HIV- infected infants have a tendency to lowered total thoracic and, perhaps, lung compliance, which may make them more susceptible to a decline in Crs after TAC. Because the subjects were studied supine and total respiratory system compliance rather than lung compliance was measured, it was not possible to differentiate the impact of TAC on the lung from its impact on total respiratory system mechanics. An alternative explanation for this finding would be that congenitally HIV-infected infants might be born with smaller airways. However, we found no differences in Rrs between HIV-infected and uninfected infants. We found no other factors such as poor nutrition which might impact on the development of the chest wall or lungs in these infants.

We did not assess sleep state or end-tidal CO₂ during the sleep induced by chloral hydrate required for the study of infant lung function. Thus, we are unable to comment on the potential impact of sleep state on the ventilation and lung volume of our infant subjects. Certainly, sleep-induced hypoventilation could result in atelectasis. However, we performed continuous monitoring of oxygen saturation by pulse oximetry from administration of the chloral hydrate until arousal was completed. We did not detect in any infant a change in oxygen saturation nor a change in respiratory rate during the duration of the infant pulmonary function study.

Finally, while the percent change in Crs and Trs induced after TAC was statistically significant, the decline in these parameters was numerically small. However, this study supports the hypothesis that TAC has a clear impact on the measurement of other parameters of infant lung function through induction of changes in chest wall and/or lung compliance.

The results of this study suggest that more careful planning of infant lung function testing is necessary to avoid artifacts or inaccuracies derived from measurement of volume-sensitive parameters after application of TAC. This is expecially true if infant lung function is to be measured preadministration and postadministration of a particular therapy. In addition, young infants and infants with conditions associated with low lung compliance are likely to be most at risk for these post-TAC changes in lung function.

Conclusions

TAC in the infant results in reduced Crs. This change in infant lung mechanics may be due to a change in lung volume, development of atelectasis, or change in the mechanical characteristics of the chest wall. Infants with abnormal lung function or at risk for abnormal lung function are likely to experience greater change in these parameters after TAC than infants with normal lung function. Therefore, when infant lung function testing includes TAC to study forced expiratory flow limitation, measurements of passive mechanics of the respiratory system should precede TAC or if repeated testing will be performed, the sequence used in measuring the various infant lung function parameters should be the same each time the testing is repeated. In addition, considering the reduction we observed in Crs after TAC, it is preferable to reserve performance of TAC as the last parameter tested.