INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

During the past 20 yr, the rapid thoracoabdominal compression (RTC) technique for measuring maximal flows at FRC (Emax_FRC) from partial expiratory flow-volume (PEFV) curves has become the most popular method of assessing peripheral airway function in infants (1, 2). Although there are limited reference data on expiratory flow in newborn infants, those that are available suggest that maximal flow at end expiration may be relatively high with respect to body or lung size (3). Such observations have been attributed to the earlier formation of airways than of alveoli during fetal development and to differences in the relative size of central and peripheral airways (i.e., dysanaptic growth) (3, 6). However, an alternative explanation could be that the dynamic increase in resting lung volume known to occur in newborn infants (9, 10) is responsible for this phenomenon (i.e., that the relatively high values of Emax_FRC in this age group could reflect differences in breathing strategy, rather than structural differences in airway caliber). It is well recognized that one of the major limitations of the RTC technique is the variability of FRC as a volume landmark. Such variability is likely to be most pronounced during the first weeks of life, when marked expiratory braking can occur (9, 11).

We became aware of the potential magnitude of this problem when studying preterm infants. Data from a 2 kg, 3-wk-old preterm infant, in whom Emax_FRC decreased from 212 to 145 ml/s during a 10-min recording period despite the use of similar jacket compression pressures, are shown in Figure 1. We attributed this fall in Emax_FRC to a change in end expiratory level (EEL) during the measurement period, which was confirmed by manually overlaying the data along the descending, flow-limited portion of the infant's flow-volume curves. This confirmed that a shift in EEL by as little as 3.5 ml could account for the observed reduction in Emax_FRC in this infant.

View larger version (19K): [in this window] [in a new window]   Figure 1.   Intrasubject variability of EmaxFRC. The two partial expiratory flow-volume curves were recorded 10 min apart at identical values of Pj. By superimposing the curves along their descending limbs, it can be seen that a shift in EEL of as little as 3.5 ml can account for a change in EmaxFRC of 67 ml/s.

Although this degree of variability is unusual, it can be problematic, especially in young infants with marked expiratory braking during the first weeks of life (9, 11). Furthermore, when studying preterm infants, we have found that it is often difficult to reproduce maximal flows once the optimal jacket pressure (P_j) (i.e., the lowest jacket pressure at which the highest flows can be attained) has been exceeded. In other words, it is sometimes impossible to obtain such high flows on subsequent occasions, even when returning to the same pressure, suggesting that marked changes in FRC, possibly related to repeated administration of thoracic compression to a compliant chest wall, may occur during the test period. Consequently, in this age group it is often necessary to perform repeated measurements with each increment of P_j if reproducible maneuvers are to be obtained, rather than rapidly increasing P_j in order to estimate optimal pressure before obtaining repeat measures, as one would in older infants.

Recent development of the raised lung volume technique for measuring forced expiratory flows from full flow-volume curves in infants (14, 15) has provided the means for examining this problem more objectively (16). The aim of this study was to assess the extent to which dynamic elevation of FRC influences values of Emax_FRC during the first month of life, by superimposing forced expiratory flow-volume curves obtained during both tidal and raised volume maneuvers.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Subjects

Infants were recruited from the Neonatal Unit at Homerton Hospital, Hackney, East London. The infants were studied while unsedated, at 0.5 to 1 h after a feeding, during behaviorally determined, natural quiet sleep (2) in room air. Respiratory measurements were obtained with the infant in the supine position while heart rate, oxygen saturation, and end-tidal CO₂ were monitored. The study was approved by the East London & City Research Ethics Committee. Informed written consent was obtained from the infants' parents, who were usually present during the measurements. Demographic details of the infants are summarized in Table 1.

Equipment and Data Acquisition

A transparent Rendell-Baker face mask (size 0; Ambu International, Bath, Avon, UK) was held over the infant's mouth and nose, and a leak free seal was created using therapeutic silicone putty (Carters, Bridgend, Mid Glamorgan, Wales). Flow was measured with a heated Hans Rudolph pneumotachograph (Model 3500; Hans Rudolph, Kansas City, MO; linearity 0 to 35 L/min) connected to a ± 0.2 kPa (2 cm H₂O) differential pressure transducer (Furness Controls Ltd., Bexhill, East Sussex, UK). Volume was obtained by digital integration of the flow signal.

PEFV curves were obtained as previously described in detail (17). Forced expiration was induced by inflating a jacket that was wrapped snugly around the infant's torso with the arms outside the jacket. The jacket extended from under the infant's axillae to the iliac crest; it consisted of a 17 × 16-cm polythene inflatable plate (Hannover, Germany) surrounded by a stiff outer fabric covering (Columbus, OH), and was rapidly inflated from a 100-L pressurized reservoir connected to the inflatable plate by rigid, large-bore (28-mm I.D.) tubing. Pressure at the airway opening (Pao) and jacket pressure (P_j) were measured with ± 5 and 10 kPa (50 and 100 cm H₂O) differential pressure transducers, respectively (Furness Controls). Flow and pressure signals were amplified and filtered above 10 Hz. Analog signals were digitized at 200 Hz (RASP, Physiologic Ltd., Newbury, Berks, UK).

Special Features for the Raised Volume Rapid Thoracoabdominal Compression Technique

The equipment used for the raised volume rapid thoracoabdominal compression (RVRTC) technique (Figure 2) was adapted from that described by Feher and coworkers (15). The pneumotachograph was attached to a mainstream capnograph (CO₂SMO, Model 7100; Novametrics Medical Systems Inc., Wallingford, CT) and a Y-piece connector; total resistance was 0.570 kPa · L¹ s at a flow of 100 ml/s. The inspiratory side of the Y-piece (three-way connector) received a constant airflow at 12 L/min via a pressure relief valve (Neopuff, RD1000; Fisher & Paykel Healthcare, Auckland, New Zealand), which was set to 2 kPa.

View larger version (20K): [in this window] [in a new window]   Figure 2.   Schematic illustration of the equipment used to generate forced expiratory flow-volume curves from raised lung volume maneuvers.

Study Protocol

The study began with measurements of Emax_FRC made with the standard RTC maneuver (17, 18). Real-time signals of flow, volume, and pressure were displayed as time based and X-Y plots. Once five to 10 regular breaths had been recorded, the jacket was inflated at end-inspiration to force expiration. P_j began at 2 kPa, and was increased in increments of 0.5 to 1 kPa until flow at FRC had reached a reproducible maximum (i.e., Emax_FRC) and higher pressures were causing a reduction of flow, indicating that apparent flow limitation had been achieved. At least three assessments of static P_j transmission were performed at end-tidal inspiration (19).

Measurements at increased lung volumes were made as previously described (16). The lowest P_j required to achieve Emax_FRC during tidal maneuvers was used during the raised volume maneuvers. To ensure that no additional dead space was presented to the infant, manual inflations began as soon the capnograph and the Y-piece were connected to the apparatus. Repeated occlusions of the expiratory side of the Y-piece, at a frequency approximating the infant's respiratory rate, resulted in passive inflations and deflations of the respiratory system. Once muscle relaxation had been achieved, as indicated by inspection of the flow and volume traces, the jacket was inflated at the end of a passive inflation to force expiration.

Data Analysis

Analog signals were digitized at 200 Hz and analyzed with a software package developed in collaboration with the Imperial College of Science, Technology and Medicine as described previously (16). Criteria for acceptability of forced expiratory maneuvers were:

A regular VT and a stable EEL for the standard RTC, or regular relaxed inflations for the raised volume RTC (as assessed by visual inspection of the time-based and flow-volume curves).

Jacket inflation begun within 100 ms of end inspiration, with a jacket inflation time of less than 100 ms and peak expiratory flow being achieved before 50% of inspired volume had been expired.

Expiration proceeding beyond the previous EEL.

A smooth flow-volume curve, without significant glottic closure or flow transients, especially during the last half of expiration, and no evidence of leaks.

Estimation of Elastic Equilibrium Volume

In an attempt to find a more reliable volume landmark with which to relate forced flows and volumes during both partial and full flow-volume curves, we calculated the elastic equilibrium volume (EEV) (i.e., the relaxed lung volume determined by the balance between the outward recoil of the chest wall and inward recoil of the lung) by extrapolating the expiratory time constant (_rs) of the passive breaths preceding each forced expiration, with EEV being taken as the volume at which zero flow crossing occurred (16) (Figure 3a). The time constant was calculated over the linear descending portion (usually the last 50 to 60% of expiration) of from two to five ensembled passive expiratory flow-volume curves preceding the forced maneuver. Results were accepted only if a time constant with an r² of at least 0.98 could be calculated over at least 45% of the expired volume toward end expiration (a portion that represented well over 75% of the duration of expiration in most of these infants). The process was repeated for each forced maneuver. The program also provided the means to overlay the flow-volume loops from a number of separate RVRTC maneuvers, and to align them according to their EEV values. Forced expiratory flows could then be calculated at any number of user-specified points above or below the EEV (16) (Figure 3b).

View larger version (17K): [in this window] [in a new window]   View larger version (22K): [in this window] [in a new window]   Figure 3.   (a) Calculation of the passive expiratory time constant (rs) by least-squares linear regression of the descending linear portion of several passive expiratory flow-volume curves preceding the forced expiratory maneuver (only one is shown for clarity). This time constant was extrapolated to zero flow to obtain the elastic equilibrium volume (EEV). (b) Alignment of forced expiratory maneuvers with respect to their EEVs. For clarity, only three maneuvers are shown, although this process could be performed with as many curves as desired. The forced expiratory flow could be calculated at any specified volume above or below the EEV. In this example, flows at EEV and at 3 ml/kg above EEV (FEEV+3) are indicated.

Overlay of Partial and Full Expiratory Flow-Volume Curves

For each infant, the three PEFV curves selected for calculation of Emax_FRC were superimposed along the descending, flow-limited portion of the raised volume curves (Figure 4). The volume above EEV at which mean FRC had been maintained during the measurement of Emax_FRC was then estimated for each infant. This was not achieved simply by measuring the distance on the volume axis after overlaying the partial and full flow-volume curves, but by numerical calculation of the volume above EEV (in ml/kg) at which the mean forced expiratory flow most closely corresponded to the mean Emax_FRC, thereby allowing an estimation of the average degree of dynamic elevation of FRC in each infant. Overlay of the partial and raised lung volume curves was done primarily as a quality control measure, to ensure that such an overlay was feasible and that flow limitation had been reached (20). Because it is possible, in the presence of expiratory braking, to generate many different profiles of flow at given volumes (especially during early expiration), such curves were excluded, together with any that were distorted in late expiration because of early inspiration.

View larger version (25K): [in this window] [in a new window]   Figure 4.   Overlay of partial and raised lung volume flow-volume curves for three infants. For clarity, only one of each type of curve is shown for each infant. Infant 1, who was wheezy, breathed out to the EEV during tidal breathing. By contrast, FRC was increased by an average of 3 ml/kg and 4.5 ml/kg, respectively, in Infants 5 and 10, who were asymptomatic at the time of testing (for infant details, see Table 1).

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Technically acceptable data for both raised volume and tidal RTC maneuvers in which the partial and full flow-volume loops could be superimposed during late expiration, including the area encompassing the end expiratory level during tidal breathing (i.e., FRC), were available for 12 of the 14 infants studied during the first 4 wk of postnatal life. Data for one infant had to be excluded because of excessive flow transients during raised volume maneuvers (16), whereas overlay of the partial and full curves for another infant was possible only during late expiration, below the FRC that was operational during tidal breathing.

Details for infants from whom acceptable data could be obtained are presented in Table 1. For clarity, only the forced expiratory flows at 2 ml/kg and 3 ml/kg body weight above EEV (i.e., F_EEV+2 and F_EEV+3, respectively) are reported in this table. Examples of the superimposed flow-volume curves from the tidal and raised volume maneuvers in three of the infants are shown in Figure 4. Infant 1, who was experiencing early wheezing following relatively prolonged oxygen requirements, was the only symptomatic infant at the time of study (Table 1). This infant had the lowest values of Emax_FRC (11 ml/s), and an FRC that coincided with EEV. By contrast, FRC was on average 3 ml/kg above EEV in Infant 5, whereas in Infant 10, FRC was even higher, at 4.5 ml/kg above EEV.

Within this group of infants, Emax_FRC ranged from 11 to 190 ml/s, with FRC being dynamically increased above EEV by 0 to 5 ml/kg (Table 1). There was a strong relationship (r² = 0.88) between these two parameters, as shown in Figure 5. Among this group of infants, the mean forced expiratory flows at 3 ml/kg above EEV (i.e., F_EEV+3) corresponded most closely to Emax_FRC (mean ± SD: 94 ± 26 ml/s, and 95 ± 52 ml/s, respectively).

View larger version (12K): [in this window] [in a new window]   Figure 5.   Relationship between EmaxFRC and the extent to which FRC was dynamically elevated above the elastic equilibrium volume (EEV).

In one infant, despite our ability to obtain three reproducible values of Emax_FRC at the optimal P_j, we noted a very high variability (ranging from 76 to 212 ml/s) during other maneuvers at the same P_j (approximately 7 kPa). By superimposing the PEFV curves for these maneuvers on those obtained at increased lung volumes, we were able to show that the range of Emax_FRC recorded in this infant could be entirely attributed to the change in EEL that had occurred during the measurement period (Figure 6).

View larger version (21K): [in this window] [in a new window]   View larger version (26K): [in this window] [in a new window]   Figure 6.   Extremes of EmaxFRC recorded in Infant 6 at identical values of Pj. By superimposing the partial flow-volume curves shown in (a) and (b) on that obtained from the same infant at raised lung volume (c), the magnitude of the shift in end-expiratory level during the recordings of the partial curves became apparent.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

By superimposing PEFV curves on those obtained with raised lung volumes in the same infant, we were able to show that the marked intrasubject changes in Emax_FRC that can occur during a single occasion of measurement, are largely attributable to changes in EEL. Furthermore, by using the EEV determined from passive expirations as a volume landmark, we could determine the extent to which elevation of FRC above EEV influences measurements of Emax_FRC in unsedated preterm infants. Within this relatively homogeneous group of preterm neonates, we found that almost 90% of the intersubject variability in Emax_FRC could be explained by the extent to which FRC was dynamically increased above the EEV at the time of measurements (r² = 0.88).

Methodology

Interpretation of the findings of this study relies on accurate determination of the EEV, which in turn depends on either complete expiration to the passive lung volume determined by the mechanical properties of the respiratory system, or on the presence of a single time constant that enables accurate extrapolation to zero flow (2). EEV was estimated by extrapolating the linear descending portion of the passive expiratory flow-volume loop. The accuracy of this approach was improved by calculating the ensembled time constant from several passive breaths prior to each forced maneuver. Results were accepted only if a time constant with an r² of at least 0.98 could be calculated over at least 45% of the expired volume toward end expiration. We found that extrapolation of this portion of the curve coincided with complete expiration to passive resting lung volume in those infants in whom this could be achieved (16).

The curvilinear appearance of the initial portion of some of the passive flow-volume curves was attributable to flow transients that occur during early expiration on release of an airway occlusion as a result of gas decompression, the magnitude of which depends at least in part on the speed at which the occlusion is released (21). This early portion of the curve, which represented a very small proportion of the total duration of expiration, was always excluded in estimation of the time constant. Ideally, the infants would have been allowed to expire fully to their passively determined lung volume after lung inflation, in order to ensure an accurate estimate of EEV. Although this may be relatively easy to achieve in older infants, it is often necessary to maintain a relatively rapid ventilatory rate and a small degree of positive end-expiratory pressure during lung inflations in young infants, in order to prevent resumption of spontaneous respiratory activity. It could be argued that the inflation volume from a preset pressure would have provided a more reliable volume landmark with which to relate the forced expiratory flows, without having to estimate EEV. However, although success has been claimed with this technique in older infants (14, 15), it is much more difficult to achieve reliable results with its use in preterm and unsedated babies (16).

The analysis software used in the study was interactive and enabled calculation of forced expiratory flows and volumes at any point throughout a breath, according to operator selection (16). For clarity, only those flows at 2 ml/kg and 3 ml/kg above EEV are reported in Table 1, since these corresponded most closely to Emax_FRC in this study. The finding that FRC was on average dynamically increased by 3 ml/kg in these infants is in keeping with previous observations of the magnitude of the volume intercept during measurement of passive respiratory mechanics in young infants (22).

Variability of Measurements

The true intrasubject variability for Emax_FRC is difficult to interpret from published data, since some authors report results as the "best" curve (i.e., the highest flow achieved at FRC) (4, 26), others report the mean of from three (17, 27) to five (28) technically satisfactory values, and still others report the mean of all technically satisfactory data (5). The results in the present study suggest that selection of the highest values of Emax_FRC may bias results toward those collected when infants are breathing at their most dynamically elevated lung volume. Provided that an infant performs enough maneuvers, it will always be possible to select three curves within 10% of one another in order to achieve apparently good intrasubject reproducibility. However, this is not necessarily representative of the infant's true Emax_FRC at the infant's most commonly assumed EEL during resting breathing. Determination of the latter remains one of the greatest challenges in assessing lung and airway function in infants, since it can affect interpretation of virtually all other measurements (1, 2).

End-expiratory lung volume is known to be highly variable in newborn infants (25), with breathing strategy being strongly influenced by the effect of sleep state on expiratory timing and diaphragmatic braking (11). Because the transition from a dynamically maintained to a relaxed end-expiratory volume in human infants is thought to occur during the latter part of the first year of life (13), problems associated with such variability of EEL should be less marked when studying older, sedated infants. Nevertheless, even in such infants, differences in the resistance and/or dead space of the measuring equipment may affect the EEL (9, 10, 29), thereby introducing a systematic bias between values of Emax_FRC collected in different laboratories.

Implications for Previously Published Data

The values of Emax_FRC found in the current study are similar to those reported previously by our department during two prior investigations of the influence of ethnicity, gender, and maternal smoking on airway function in preterm infants (17, 30). In both of these studies, a number of infants were found to have unexpectedly high values of Emax_FRC for their body size. Similar observations were reported by Tepper and colleagues (27), who excluded the youngest infants from their regression equations of Emax_FRC, limiting their predictions to infants above 1 mo of age. When reviewing the literature, Beardsmore and coworkers (5) likened the shape of the relationship between Emax_FRC and body length to a "hockey stick," again emphasising the relatively high values of Emax_FRC in newborns as compared with older infants and children. In the past, the tendency toward enhanced specific airway conductance (31) and forced expiratory flows (3, 4) in the very young has been attributed to the earlier formation of airways than of alveoli, and to developmental differences in the relative size of peripheral and central airways during fetal and early infant development (7, 27, 31). The results in the present study suggest that this phenomenon may be largely attributable to differences in breathing strategy, rather than to airway structure per se.

A complex relationship exists between lung volume, vagal control of breathing (including the extent to which expiratory flow is modulated), and respiratory mechanics (11, 24), such that although there may be considerable variability of breathing strategy and end-expiratory level in healthy infants, a much more regular respiratory pattern, with minimal expiratory braking is usually observed among those with airway disease. Infants with airway narrowing and a prolonged expiratory time constant will not need to modulate their expiratory flow to maintain an adequate lung volume (7), and will tend to expire more passively toward their elastic equilibrium volume (Figure 4, Infant 1). Thus, although caution is needed in interpreting values of Emax_FRC among healthy neonates, the presence of markedly diminished flows is still likely to indicate a reduced airway caliber in those with airway disease.

The true impact of changes in lung volume during assessments of airway function has not always been appreciated. It has recently been suggested that in infants with tracheomalacia, the observed increase in Emax_FRC during treatment with continuous positive airway pressure (CPAP) is due to increased airway diameter and decreased collapsing pressure at the flow-limiting segment of the airway (32). However, because lung volume was not measured, its contribution to the observed changes in Emax_FRC could not be determined. By contrast, when Tepper and coworkers assessed the effect of CPAP through use of the raised volume technique (33), they showed that although flows at FRC did increase with CPAP, flows at the same lung volume were similar for the three levels of CPAP imposed. This is, in fact, confirmed by close inspection of the illustrations from the study by Panitch and associates (32), which show that overlay of the four curves at different end expiratory levels would indeed be possible.

Do these findings therefore invalidate measurements of Emax_FRC during the neonatal period? Not necessarily, since, as shown in Table 1, the ranking of infants according to their forced expiratory flows was similar in the present study whether Emax_FRC or F_EEV+3 was used. The only infant to have clinical symptoms of wheezing and impaired airway function also had the lowest expiratory flows, whether these were expressed as Emax_FRC or F_EEV+3 (Table 1). Similarly, the highest flows according to either parameter were found in an infant recovering from respiratory distress syndrome, in which increased recoil pressure in the presence of normal airway caliber could have contributed to the results. Previous investigations have shown that it is not necessary to initiate RTC from increased lung volumes in order to make reliable flow comparisons (34). However, the variability of the end-expiratory level will weaken the association between Emax_FRC and impaired peripheral airway function, thereby necessitating a much larger sample size to answer clinical or epidemiologic questions than would be needed if forced flows were related to a more reliable volume landmark. The potential value of superimposing the descending limb of the forced expiratory segment of PEFV curves obtained at different jacket pressures, as a means of demonstrating flow limitation, has recently been proposed (20). Some combination of partial and full forced expiratory maneuvers may prove to be the ideal means of obtaining reproducible and representative data as quickly and simply as possible in young and unsedated infants.

In conclusion, it would appear that in studying preterm infants during the first few weeks of life, caution should be exercised when using Emax_FRC. Whether our findings are equally applicable to older, sedated infants needs to be determined. Al-though, as shown in numerous previous studies (1), Emax_FRC can provide a very useful means of identifying peripheral airway obstruction in infants, the marked intra- and intersubject variability will necessitate measuring large numbers of infants if conclusive results are to be obtained in epidemiologic studies or clinical trials. More meaningful within- and between- infant comparisons of peripheral airway function may be obtained by calculating forced expiratory flows at a fixed interval (e.g., 3 ml/kg) above EEV, rather than at the FRC that is operational at the time of measurement. Further studies are required to assess the relative sensitivity and specificity of the various parameters that can be derived from raised lung volume maneuvers, in order to select the most suitable outcome measures for use in investigations of factors influencing airway development during early life.