Monograph Forced Oscillation Technique

CHAPTER 5
Forced oscillation technique and impulse oscillometry

H.J. Smith*, P. Reinhold#, M.D. Goldman} *Research in Respiratory Diagnostics, Berlin, Germany. #Friedrich-Loefer-Institute, Jena, Germany. } David Geffen School of Medicine, University of California, Los Angeles, USA. Correspondence: H.J. Smith, Research in Respiratory Diagnostics, Bahrendorfer Str. 3, 12555 Berlin, Germany.
Conventional methods of lung function testing provide measurements obtained during specic respiratory actions of the subject. In contrast, the forced oscillation technique (FOT) determines breathing mechanics by superimposing small external pressure signals on the spontaneous breathing of the subject [1]. FOT is indicated as a diagnostic method to obtain reliable differentiated tidal breathing analysis. Because FOT is performed without closure of a valve connected to the mouthpiece, and without maximal or forced respiratory manoeuvres, it is unlikely that FOT itself will alter airways smooth muscle tone [2]. FOT utilises the external applied pressure signals and their resultant ows to determine lung mechanical parameters. These pressureow relationships are largely distinct from the natural pattern of individual respiratory ows, so that measured FOT results are, for the most part, independent of the underlying respiratory pattern. Therefore, oscillometry minimises demands on the patient and requires only passive cooperation of the subject: maintenance of an airtight seal of the lips around a mouthpiece and breathing normally through the measuring system with a nose-clip occluding the nares. Potential applications of oscillometry include paediatric, adult and geriatric populations, comprising diagnostic clinical testing, monitoring of therapeutic regimens, and epidemiological evaluations, independent of severity of lung disease. Oscillometry is also applicable to veterinary medicine. The last two main sections Relevance of oscillometry in clinical practice and Oscillometry in the clinical pulmonary laboratory emphasise clinical aspects of application and interpretation of FOT rather than methodological details and technological solutions, which are discussed in the two sections that follow immediately below. Clinical application of FOT does not require mastery of the mathematical infrastructure of the technical methodology, and readers interested in the clinical use of FOT may nd it more useful to begin with these clinical sections and refer subsequently to methodological and technological details.
Methodology of impulse oscillation technique

The mechanical basis of oscillometry concerns use of external forcing signals, which may be mono- or multifrequency, and applied either continuously or in a time-discrete manner. The impulse oscillation technique is characterised by use of an impulse-shaped, time-discrete external forcing signal.
Eur Respir Mon, 2005, 31, 72105. Printed in UK - all rights reserved. Copyright ERS Journals Ltd 2005.
72
OSCILLOMETRY: FOT AND IOS
The most useful aspect of applying FOT pressure pulses rather than pseudo-random noise (PRN) is improved time resolution of the measurement. The impulse oscillation technique allows measurement of up to 10 impedance spectra per second. This permits a useful analysis of intra-breath variation in impedance, comparable to that obtained with mono-frequency applications. However, a disadvantage of such high impulse rates is the inability to record longer respiratory time constants that may be more informative concerning respiratory abnormalities. For this reason, the common application of impulse oscillation utilises recordings of 5 impedance spectra (5 impulses) per second. An additional benet of impulse oscillation is the simplicity of the hardware needed to generate the forced signals, allowing smaller, more efcient electronic and mechanical structures with minimal power loss. A unique aspect of applying pulses of pressure to the respiratory system is the fact that the entire energy of all applied pressure harmonics is applied within a very short time period. This causes a higher impact to the respiratory system compared with sinusoidal or PRN applied pressures, and may be perceived by some patients as a slightly unpleasant respiratory sensation during measurement.
Peculiarities of aperiodic waveforms

The impulse oscillation method applies aperiodic waveforms using an impulse generator that produces pressure pulses of limited magnitude and 3040 ms duration. These pulses dene specic amplitudes and phases of the inherent sinusoidal components. The time-course of such practical pressure pulses applied to the respiratory system is not a true Dirac-impulse, dened as having virtually innite magnitude and innitesimal time duration, which would provide a continuous spectrum of frequencies with the same amplitude. Thus, the terms "impulse-shaped" oscillations and "impulse pressures" are used to indicate realistic practical pressure pulses rather than mathematically dened impulses. The short duration of the impulse-shaped waveform itself provides linearity of pressure and ow signals in the face of within-breath dynamic changes in the respiratory system. In contrast, the longer time needed with PRN to embed a range of periodic functions decreases time resolution, resulting in increased noise of calculated impedance related to any time-dependence of dynamic changes in the respiratory system. The characteristic feature of any aperiodic waveform is the resulting continuous spectrum after transformation of its time course into the frequency domain, using a Fourier integral rather than a Fourier series, in the fast Fourier transform (FFT). Thus, the advantage of continuous spectra is particularly important in abnormal respiratory systems with regional nonhomogeneities (g. 1), where resistance, reactance and coherence spectra may manifest deviations from the normally smooth and uniformly continuous spectral courses. In contrast, spectra resulting from FFT analyses of periodic multifrequency forcing such as PRN [3] are discontinuous. Discrete values of impedance are obtained with a frequency resolution determined by the frequencies of the included sinusoidal components. As a result, the course of such discrete spectra often may require postprocessing to smooth the PRN spectra [3]. To improve interpretation of discrete PRN spectra, it is common to approximate the overall spectral range by smoothing with linear, quadratic or logarithmic functions. However, such smoothing inherently diminishes information contained in characteristic peaks and plateaus of impedance that may otherwise provide insight into superimposed parallel resonance phenomena, e.g. related to upper airway constriction. Impulse power spectra of pressure and ow generated by the impulse oscillometry
73
H.J. SMITH ET AL.
Rrs 0.5 0.4 kPasL-1 0.3 0.2
Xrs 0.3 0.2 0.1 0
Coherence g2 () Regional nonhomogeneity Resistance Rrs ()
g2 1.0 0.9 0.8 0.7 0.6 0.5
res Reactance Xrs() Reactance area (AX)
0.1 -0.1 0 -0.2
10
15 20 Frequency Hz
25
30
35
Fig. 1. Representative data for spectra of respiratory resistance (Rrs(f); ), reactance (Xrs(f); - - -) and coherence (c2(f); - - ) are plotted, between 3 and 35 Hz, for a normal adult, during forced oscillation using pulse-shaped forcing generated by the impulse oscillometry system. Resonant frequency (fres) is shown where the Xrs(f) tracing crosses zero. The shaded area, below zero Xrs and above Xrs(f) tracing between 5 Hz and fres, is the integral of Xrs(f) from 5 Hz to fres, and is designated the reactance area (AX). Regional nonhomogeneities may manifest deviations from the normally smooth and uniformly continuous spectral courses.
system (IOS) are shown in gure 2 over a frequency range of 0.135 Hz. The energy distribution provides practical assessment of low (5 Hz), as well as high (w20 Hz) frequency ranges, with decreased amplitude at higher frequencies to minimise nonlinearities due to acceleration of the moving air column [4]. Enhanced amplitudes at the
5 0 Attenuation dB -5 -10 -15 -20 -25
10
15 20 Frequency Hz
25
30
35
Fig. 2. Power spectra of ow (), and pressure (----), for the discrete pulse-shaped forcing signal generated by the impulse oscillometry system. Spectra plotted at frequencies 0.135 Hz. Pressure and ow power are highest at 320 Hz. Less power is needed at frequencies w20 Hz, because "competing" higher harmonics of the patients respiratory ow are very small at these frequencies.
74
lower frequencies limit the inuence of higher harmonics of spontaneous breathing frequencies.
Impulse oscillometry system

The IOS measuring-head (g. 3) is functionally similar to PRN systems designed for the determination of input impedance [1, 512]. The characteristic feature of the IOS [13] is the generation of recurrent aperiodic impulse-shaped forcing signals of alternating direction. Flow is measured by a Lilly-type heated screen pneumotachograph with a resistance of 36 Pa?s?L-1, providing a common-mode rejection ratio ofw60 dB up to 50 Hz [12, 14, 15] for the combination of pneumotachograph and ow transducer system. At ow rates v15 L?s-1 the heated pneumotachograph is linear within 2%. The proximal side of the pneumotachograph is connected to a pressure transducer. To guarantee suppression of technical inuences and to avoid phase differences, both pressure and ow channels use matched transducers of the same type, SensorTechnics SLP 004D [16]. Pressure and ow signals are sampled at a frequency of 200 Hz and digitally converted by a 12-bit analogue-to-digital converter. Analogue anti-aliasing low-pass ltering is realised by a fourth-order Bessel lter at a border frequency of 50 Hz, providing a damping at Nyquist frequency of y75 dB. Tuning of the IOS impulse generator involves both volume displacement of the loudspeaker membrane and magnitude of the terminating resistor. The terminal resistor
RS 232 to computer DSP Loudspeaker
Impulse generator Y-adapter
Terminal resistor
Metal screen
w flo lse flow pu y Im ator ir sp Re
Pneumotachograph (heated)
Flow transducer Mouthpiece Pressure transducer

Fig. 3. Schematic diagram of the impulse oscillometry system measuring-head and connectors: loudspeaker enclosure at top, connected to a y-adaptor at one upper arm, an exit for ow with terminal resistor at the second upper arm and a lower arm connected to the pneumotachograph. A mouthpiece is connected to the open side of the pneumotachograph. Pulsatile ows generated by the loudspeaker are shown as a lightly shaded thick line, part of which exits through the terminal resistor and part of which ows through the pneumotachograph and mouthpiece. The patients normal respiratory ow is shown as a shaded thick line from the mouthpiece though the Y-connector, exiting through the terminal resistor. The resistance of the terminal resistor is 0.1 kPa?s?L-1 and the deadspace of the Y-connector/pneumotachograph/mouthpiece is y70 mL.
75
H.J. SMITH ET AL.
provides a low-impedance pathway for normal respiratory ow, which, at the same time, is high enough to minimise loss of energy of superimposed impulses so that sufcient impulse pressure is transmitted into the respiratory tract. Both components determine the linear working range of the unit and range of input impedances, which can be measured to maintain international recommendations [16, 17]. A terminal resistor of y0.1 Pa?s?L-1, in combination with a volume displacement of 40 mL, which is accelerated by the loudspeaker membrane in v40 ms, results in maximal peak-to-peak impulse pressures of 0.3 kPa and minimises interference of underlying higher harmonics of respiratory frequency that contribute "noise" to the oscillatory pressure and ow signal [10]. International recommendations for electromechanical performance are maintained in the IOS by use of advanced transducer technology and global mean spectral data derived from IOS are generally comparable to those obtained by the pseudo-random noise ` method of Landser [6], Delecourt et al. [8] and Skloot et al. [18]. The measurement is performed as follows: while the subject spontaneously breathes ambient air via the tubing between mouthpiece and terminating resistor, the loudspeaker generator transmits brief pressure impulses via Y-adapter, pneumotachograph and mouthpiece into the respiratory tract; pneumotachograph and pressure transducer register the composite signals containing breathing activities and the forcing impulse signal for further processing.
Further processing of digitised impulse data. Flow and pressure channels contain both
the underlying respiratory system ow and pressure, and the superimposed forced oscillation signals. By denition, respiratory input impedance is the transfer function or ratio of effective pressure (Prs) and ow (V9rs), derived from the superimposed forced oscillations, after being discriminated from underlying respiratory pressure and ow and their harmonics. In the mathematical sense, all components are considered "complex", characterised by modulus and phase.
Prs (f ) 1 Multifrequency input impedance Zrs(f )~ | f0 < f f maxg V 0 rs (f ) | Discrimination of superimposed forced oscillations from the underlying respiratory pressure and ow in the IOS is focused on individual impulses, based on pressure and ow sampling intervals that include both the impulse stimulus and the respiratory system reaction to the impulse. Figure 4a gives an example of such a sampling interval for ow. A "baseline" straight line segment is inserted between the start- and end-points of separate sample segments of both pressure and ow. The baseline is a simple linear approximation of the underlying respiratory ow and pressure throughout the single impulse excitation. Baseline approximation has proved to be a useful and reliable technique to decrease the respiratory component of the composite signals for pressure and ow. Spline reconstruction, sinusoidal approximation of underlying respiratory pressure and ow or high-pass digital ltering were not as useful. Baseline correction and offset elimination, as can be seen in gure 4b, allow rectangular windowing prior to the FFT to effectively reduce spectral leakage and improve the signal-to-noise ratio. Resolution of calculated pressure and ow spectra is increased by adding numerical zero values to the real sampling points of the corrected primary data segment. This procedure allows formation of exactly 2n samples, compatible with FFT requirements. Choice of sampling interval as well as addition of zero values to this interval allow
76
a) 1.00 Amplitude of primary flow Ls-1 0.75 0.50 0.25 0.00
Segment 160 ms
Approximated baseline Pulserate 5s-1 200 ms
-0.25 -0.50 0.0 0.1 0.2
b) Corrected impulse flow Ls-1
0.4
0.3 0.4 Recording time s
0.5
0.6 150
0.7
0.2
Resulting impulse pressure
75
Reactive response of respiratory system 0.0 Duration of generator impulse 0 30 Baseline Impulse flow -75 0
-0.2
60 90 Sampling time ms
120
150
Fig. 4. a) Primary ow recording. Note that the primary ow includes the patients expiratory respiratory ow with pulses of "loudspeaker ow" superimposed. The dash-dot line shows the linearly approximated respiratory ow if there were no superimposed pulsatile ow from the loudspeaker included. By referring all pulsatile ow and with a similar process utilised for the continuous pressure tracing to the approximated baseline, the ow and pressure related purely to the pulse generated by the loudspeaker, with the patients respiratory ow and pressure subtracted from the total pressure and ow signals, may be derived. b) Corrected impulse tracings of ow () and pressure (----), derived from baseline correction shown in gure 4a, as prepared for input into the fast Fourier transform. Duration of impulse corresponds to actual movement time of the loudspeaker diaphragm (y40 ms in duration). Initial ow response of the respiratory system begins almost immediately. After loudspeaker movement has ended, the respiratory system continues to respond in a "reactive" manner, with pressure reaching its peak and then declining, while ow falls below baseline, and then gradually returns to baseline.
adjustment of spectra concerning low-end border frequency as well as numerical resolution. To further improve quality of calculated impedance data, impulses that do not fulll dened reliability criteria are rejected. The critical segments of respiration are the phase transitions between inspiration and expiration. At these zero-crossings of pressure and ow, gradients of pressure and ow versus time are maximal, and the following principles are implemented to establish reliability. Slopes of baseline corrections for pressure or ow w0.7 indicate dominance of the underlying respiratory pressureow pattern, and the
77
Corrected impulse pressure Pa
H.J. SMITH ET AL.
impulse is rejected because separation of the superimposed impulse pressure and ow is not reliable. Absolute peak ow within the impulse segment must exceed 0.02 L?s-1. Pulses with less ow are rejected because of very small ow values in the denominator of the input impedance Equation (1) [19] and the resulting mathematical errors. Finally, impulses that yield negative resistance values at any frequency after FFT are rejected.
Impulse rate and sampling interval. Impulse rate and selected sampling interval effects
on calculated impedance have been evaluated in vitro as well as in calves of comparable size to adult humans [20, 21]. No signicant effect of impulse rate was observed between 1 and 5 impulses per second. In contrast, different sampling durations led to signicantly different low-frequency respiratory system reactance (Xrs) results. Using 16 sampling points (80 ms) for data analysis, no useful information was obtained for Xrs 5 Hz. Use of 32 sampling points (160 ms) per impulse for data analysis provided useful Xrs5 data. However, sampling durations of 320 ms did not improve impedance results and coherence c2 5 Hz was signicantly decreased, consistent with deterioration of calculated impedance quality due to interactions with spontaneous breathing signals. These ndings underlie the current recommendation to use sampling intervals of 32 sampling points per impulse, corresponding to an optimised impulse rate of either 3 or 5 impulses per second.
Coherence. The coherence function is dened as the square of the cross-power spectrum
divided by the product of the auto-spectra of pressure and ow at any forcing frequency. Ranging between 0 and 1, it is a measure of the available linearity [22].
jGV 0 P (f )j2 f0 < f f maxg 2 GV 0 V 0 (f ):GPP (f ) ` E Landser et al. [6] found that when the coherence c2 i0.95, PRN impedance data show a coefcient of variation CV%v10%. Subsequently, coherence value thresholds of 0.9 or 0.95 have been widely used to accept FOT data. However, if methods other than PRN forcing input signals or data acquisition are used, this original threshold value of coherence is not applicable. Miller and Pimmel [23] showed that estimated variance of calculated impedance is a function of coherence and the number of estimates averaged. Use of pseudo-random noise techniques commonly includes three measures of impedance of 16 s each [5]. Three replicate measures require coherence c2 w0.9 to yield an estimated standard error of 10%. In contrast, IOS commonly includes the average of w100 separate FFT analyses and, accordingly, requires less perfect coherence for the average data to provide an estimated standard error of v10%. For clinical purposes it is recommended to use a coherence value of c2 i0.6 at 5 Hz for the acceptance threshold, provided that at least 100 FFT analyses are averaged. Coherence improves as a function of oscillatory frequency to c2 i0.9 at 10 Hz and higher frequencies (g. 1). c2 (f )~
Interpretation of oscillation mechanics

Monofrequency oscillations provide a measure of total respiratory impedance (Zrs) that includes airway resistance, and elastic and inertive behaviour of lungs and chest wall at one oscillation frequency. In contrast, multifrequency oscillation methods, such as pseudo-random noise or impulse oscillation provide measures of respiratory mechanical properties in terms of Zrs, as a function of frequency (f), allowing the recognition of characteristic respiratory responses at different oscillation frequencies.
78
Zrs has been described in engineering terms by so-called "real" and "imaginary" components. In medical use, preferred terms are resistance (Rrs) and reactance (Xrs), respectively.
Zrs(f )~Rrs(f )zjXrs(f ) f0 < f f maxg 3 | In contrast to Rrs, concepts of Xrs are not yet widely appreciated, largely because of greater complexity of reactive parameters, as well as their numerical characteristics. The present discussion emphasises the importance of consideration of both Xrs as well as Rrs for interpretation of respiratory mechanical properties. Most commonly, the oscillation frequency scale utilised for multifrequency oscillation methods includes about 530 Hz [7, 17, 24]. Oscillation frequenciesv5 Hz are affected by higher harmonics of underlying natural respiratory frequency [25]. Higher oscillation frequencies are increasingly affected by "shunt" properties of the upper airways [6, 13, 26] and capacitive impact of face masks [27] when used.
Respiratory resistance
The resistive component of respiratory impedance, Rrs, includes proximal and distal airways (central and peripheral), lung tissue and chest wall resistance. Normally, central resistance dominates, depending on airway calibre and the surface of the airway walls, while lung tissue and chest wall resistance are usually negligible [7, 28]. Rrs may be considered within normal limits if Rrs at 5 Hz (Rrs5) is within 1.64 sd of the predicted value. Rrs5 values between 1.64 and 2 sd above predicted may be considered minor, w2 sd moderate and w4 sd above predicted severe obstruction. In previous reports the calculation of per cent predicted has been used. Rrs5 values that did not exceed 150% predicted, dened in bronchial challenge comparing changes in Rrs5 to a 20% decrease in forced expiratory volume in one second (FEV1) and 50% increase in airway resistance (Raw), were considered within normal limits [2931]. However, it is now recognised that a 20% decrease in FEV1 is a substantial abnormality, and normal limits for Rrs might be more protably dened in their own right, without requiring a specic relationship to arbitrary spirometric criteria. In healthy subjects, Rrs is almost independent of oscillation frequency [7, 32, 33], but may increase slightly at higher frequencies due to the upper airways shunt effect [6, 26]. When proximal (central) or distal (peripheral) airway obstruction occurs, Rrs5 is increased above normal values. The site of airway obstruction is inferred from the pattern of Rrs, as a function of oscillation frequency, adjusting as necessary for subject age [2, 25, 3440]. Proximal airways obstruction elevates Rrs evenly independent of oscillation frequency [32]. In distal airways obstruction, Rrs is highest at low oscillation frequencies and falls with increasing frequency. This negative frequency-dependence of Rrs has been explained in terms of intrapulmonary gas ow redistribution, due either to peripheral pulmonary nonhomogeneities or to changes in peripheral elastic reactive properties [6, 8, 34, 35]. As peripheral resistance increases, Rrs becomes more frequency dependent [26, 3638]. Frequency dependence of Rrs may be a normal nding in small children [39, 40] and may be greater than in adults in the presence of peripheral airow obstruction [8].
Respiratory reactance
The reactive component of respiratory impedance, Xrs incorporates the mass-inertive forces of the moving air column in the conducting airways, expressed in the term inertance (I) and the elastic properties of lung periphery, expressed in the term
79
H.J. SMITH ET AL.
capacitance (Ca).
1 v~2:p:f f0 < f f maxg 4 Xrs(f )~v:I{ : v Ca Most importantly, in medicine, it is emphasised that respiratory Ca is not identical to compliance. The component of Xrs associated with Ca is dened to be negative in sign. It is most prominent at low frequencies. In contrast, the component of Xrs associated with inertance is always positive in sign and dominates at higher frequencies. Thus, interpretation of Xrs is primarily inuenced by the oscillation frequency range under consideration. Low frequency capacitive Xrs essentially expresses the ability of the respiratory tract to store capacitive energy, primarily resident in the lung periphery. In both brosis and emphysema this ability is reduced: in brosis because of the stiffness of the lung; in emphysema because of hyperination and loss of lung elastic recoil. Distal capacitive reactance at 5 Hz (Xrs5) manifests increasingly negative values either in restriction or in hyperination. Thus, Xrs5 characterises the lung periphery, but is nonspecic as to the type of limitation. Additional information is needed to differentiate peripheral obstruction from peripheral restriction. This is not usually problematic in clinical practice. Capacitive and inertive elements have been modelled by a number of authors [2]. Many simplications have been attempted that include serial and parallel circuit elements, to dene an approximation of a normal Xrs spectrum with a zero-crossing exactly at resonant frequency and positive slope throughout. The frequency range utilised in multifrequency oscillometric forcing signals should allow for determination of the serial resonance of the respiratory system under investigation [5].
Resonant frequency. Resonant frequency (fres) is dened as the point at which the
magnitudes of capacitive and inertive reactance are equal. 1 v0 :I~ : v0 ~2:p:f res v0 Ca
1 p 6 :p: I :Ca 2 Because fres can vary over a considerable range and, thereby, appear in close proximity to oscillation frequencies dominated by either capacitative or inertive properties, this parameter should not be directly interpreted in terms of a particular mechanical property of the respiratory system. However, it is a convenient marker to separate low-frequency from high-frequency Xrs. Thus, low-frequency capacitive Xrs is dominant at oscillation frequencies below fres, while high-frequency inertive Xrs is dominant at frequencies above fres. In normal adults, fres is usually 712 Hz [33]. In healthy young children, fres is larger than in adults and increases with decreasing age. In respiratory disease, both obstructive and restrictive impairments of the distal respiratory tract cause fres to be increased above normal [7, 26, 36]. The relevance of fres is normally revealed in within-individual trends over time during bronchial or therapeutic challenge. f res~
Capacitive spectral range of reactance. Thoraco-pulmonary elasticity is commonly

viewed as a static property, and normally is investigated in the absence of resistive or inertive mechanical losses. However, in oscillometry, capacitive Xrs is believed to comprise useful information concerning peripheral airways mechanical properties. In
80
practice, determination of peripheral capacitive Xrs is determined at the lowest frequency that is not highly interfered with by fundamental respiratory frequency and its harmonics [41]. In the IOS method, Xrs5 is commonly utilised. In children breathing at high respiratory frequencies, reactance at 10 Hz may be useful [42] in following bronchial and/ or therapeutic challenge. Interpretation of Xrs5, is clearly different from that of conventional lung function test parameters and, in particular, lung compliance. One striking feature of Xrs5 is its negative value. The minus sign is derived from a general denition in natural sciences to differentiate elastic properties from moments of inertia, the latter of which is always positive. Therefore, the minus sign simply conrms a relationship with elastic properties. Denition of abnormality has previously been based on increased negative readings related to expected normal values of Xrs5. A differencew0.15 kPa?s?L-1 is most denitely agreed to imply abnormal lung function, independent of Rrs, although abnormal lung function may be present with smaller differences.
Capacitance versus dynamic pulmonary compliance. Dynamic pulmonary compliance

is derived from the relationship between oesophageal pressure and changes in lung volume [43]. In contrast, oscillometry assesses the elastic properties of the respiratory system from the out-of-phase relationship of simultaneously recorded transrespiratory pressure (Prs) and central ow signals (V9rs), measured from superimposed oscillations and transformed into Xrs. Therefore, the term capacitance, Ca, an equivalent of capacitive phase information between the primary signals Prs and V9rs should be used. The frequency range for such measures is always below fres. In pulmonary brosis, dynamic lung compliance (CLdyn) is decreased below normal. In a similar manner, oscillometry yields a decreased estimate of Ca, due to negative displacement of low frequency Xrs. Both dynamic lung compliance and Ca reect elastic limitation and they trend together. Previous oscillometry studies have reported less sensitivity than dynamic lung compliance in the early stages of restrictive disease [7, 17]. In contrast, pulmonary hyperination is associated with loss of lung elastic recoil and increased CLdyn. Because of the loss of lung elastic recoil, peripheral airways are not supported externally by lung recoil and the resultant partial peripheral airway obstruction prevents the applied oscillometric signals from reaching peripheral compliant areas. In this way, loss of lung elastic recoil indirectly causes a decrease in Ca, with associated increased negative magnitude of low frequency Xrs, [26, 36]. Indeed, low frequency Xrs is particularly sensitive to pulmonary hyperination, and while Rrs5 may be nearly normal or only moderately increased, Xrs5 is highly abnormal [44].
Reactance area. The index designated reactance area (AX) is a quantitative index of total
respiratory reactance Xrs at all frequencies between 5 Hz and fres. The integration of these negative values of Xrs [45] creates an area between the reactance zero axis and Xrs, providing an integrative function to include changes in the magnitude of low-frequency reactance Xrs, changes in fres and changes in the curvature of the Xrs(f)-tracing. It is represented graphically in gure 1 as the area under the zero axis of the reactance graph above the Xrs(f) tracing.
f res
AX ~
5
Xrs(f ):df
This integrative index provides a single quantity that reects changes in the degree of peripheral airway obstruction and correlates closely with frequency dependence of resistance [18].
81
H.J. SMITH ET AL.
Inertive spectral range of reactance. During resting breathing at normal respiratory

frequencies, transrespiratory pressure is dissipated in resistive and elastic losses, whereas inertive pressure losses are negligible [46]. In contrast, during forced oscillation, when oscillatory frequencies are more than 10-times greater than normal respiratory frequency, inertance (I) contributes signicantly to dissipation of the externally applied pressures [16, 46]. As noted above, the inertive spectral range of Xrs is at frequencies above fres. These frequencies reect mechanical properties of the proximal conducting airways. However, specic clinical interpretation of Xrs in this range is limited due to the wide variety of inuences that may appear in the upper respiratory tract and resultant resonant effects that may change Xrs unpredictably. Changes in central airway calibre correlate more strongly with resistive parameters and are less well represented in the inertive spectral range. Finally, it should be noted that oscillometric reactance occasionally demonstrates a low-to-mid frequency plateau in the Xrs (f) tracing, which is suggestive of possible upper airway obstruction [4750].
Variability of oscillometric parameters

The majority of oscillometric parameters can usefully be assessed using the coefcient of variation (CV%). Short-term variability should not exceed 10%, for magnitude of Zrs and Rrs at frequencies i5 Hz [17]. Variability of Xrs is larger, because of physiological and numerical characteristics. Xrs can be positive or negative and is commonly close to zero. As a result, the calculation of CV% is not suitable to estimate the variability of most of the Xrs values. Therefore, it is recommended to estimate variability of Rrs and Xrs using standard deviation, 95% percentiles for normally distributed values or calculating the absolute difference (range) between minimum and maximum of the Xrs parameters [20, 51].
Methodological versus biological variability. Low methodological variability in

measures of impedance (Zrs) has been shown using physical models with rather high reproducibility [20]. Biological variability is much more complex and incorporates intra-breath and intraand inter-subject variability. Intra-subject variability of consecutive measurements within a specied time period, including circadian variability, day-to-day variability and variability associated with changes in diseased airways has been reported previously [52 56]. Intra-breath variability is of specic physiological interest. Commonly, Zrs is determined as an average over a number of consecutive breathing cycles within 15 30 s. However, ow- and volume-dependence of Rrs and Xrs within each breathing cycle may be apparent during both inspiration and expiration, and have been shown in a number of investigations to reect specic pathophysiological characteristics [5761].
Special application of impulse oscillometry system to animals

Application of the impulse oscillation technique has been described in different animal species. The standard IOS device, originally developed for humans, has been validated carefully in calves aged up to 6 months and weighing 35150 kg [20, 27, 62] and in pigs [63, 64]. For large animals, such as horses or adult cattle, a specially designed IOS unit has been developed, which is capable of analysing larger ow and volume characteristics. While no data are available in adult bovines, methodological validation and clinical application of IOS in horses has been reported [6567] and is still in progress.
82
Oscillometry in spontaneously breathing conscious animals requires the imposition of applied pressure signals via a rigid mask. Without correcting measured impedance for the facemask, the limit of frequencies that can be clinically analysed is v15 Hz. Higher frequencies are substantially inuenced by the capacitance of the facemask itself [27]. The most useful frequency range for clinical evaluation of respiratory impedance differs in different species, primarily dependent on animal size. The lower the specic frequency is, the more sensitive the measurements to the periphery of the respiratory system are. For example, while the resonant frequency is between 512 Hz in calves, depending on body weight, it isv5 Hz in horses. Accordingly, frequencies that reect the lung periphery in horses are lower than in calves. In agreement with results in human medicine, peripheral airway obstruction is characterised by a marked increase in magnitude of low frequency respiratory reactance (|Xrs|) and in fres. In addition, the resistance spectrum of Rrs shows increased negative frequency dependence and increase in low frequency Rrs (v5 Hz). Upper airway narrowing is characterised by a parallel increase in Rrs at all frequencies with no change in the frequency dependence of Rrs. No signicant changes occur in reactance with upper (large) airway narrowing.
Relevance of oscillometry in clinical practice

This section offers a perspective on clinical use of the FOT method of determining Zrs that may be summarised as follows: 1) FOT provides useful clinical information that prominently includes functional assessment of small, peripheral airway behaviour beyond that available from commonly used pulmonary function tests (PFT). 2) Because of its sensitivity to peripheral airway function, FOT in its own right, apart from other PFT results, provides useful guidance in clinical patient management. 3) The prominence of peripheral airway functional assessment provided by FOT derives both from Xrs as well as Rrs. 4) The importance of Xrs is amplied by recognition of different Xrs-characteristics at low, i.e. below resonant frequency (vfres), and high (wfres) oscillation frequencies. The last issue is considered in detail in the foregoing technological sections. Briey, it is noted here that original technical descriptions of FOT included calculation of the magnitude (|Zrs|) and phase (Q) of Zrs, i.e. polar coordinates [1, 5]. This engineering description gave way to clinical research descriptions of Rrs and Xrs in Cartesian coordinates. As noted in the technology section, Xrs relates to peripheral airway properties at oscillation frequencies vfres, and to central conducting airways at frequencies wfres. The use of the magnitude of Xrs (|Xrs|) rather than Xrs itself in algebraic manipulation of reactance data is emphasised, because this increases sensitivity to changes in peripheral airway mechanical properties, as demonstrated later in this section in reference to previously reported studies. The section that follows includes discussion of monofrequency, pseudo-random noise and pulse-shaped pressure oscillations, as these three methods are currently in common use. The general principles apply to all methods of FOT for the most part where issues concern specic use of multifrequency rather than monofrequency or IOS rather than PRN this difference is stated explicitly. Previous published reviews have discussed theoretical and modeling aspects of FOT [2, 45, 68]. The present discussion does not include an exhaustive review of clinical research investigations from the Barcelona group [6973], Leuven [6, 29, 32, 36, 39, 7478], London [7, 7981], Paris [8, 8286] and
83
H.J. SMITH ET AL.
Vandoeuvre-les-Nancy [25, 8791], which have all helped to provide the essential infrastructure for clinical application of oscillometry. Instead, the authors focus on published studies in relation to current work that permit practical establishment of oscillometry in the routine clinical pulmonary function laboratory. Furthermore, because the authors have worked more intimately with IOS, illustrative examples of current work with this technology are included. The relative advantages of each type of FOT may be summarised by noting that the simplest form for clinical practitioners is monofrequency sinusoidal pressure application [2, 68]. Measurement of Rrs with this method is applicable to patients with sleep apnoea and those using continuous positive airway pressure or mechanical ventilation. Multifrequency FOT provides further characterisation of respiratory mechanics, including variation of Rrs and Xrs with oscillation frequency. PRN FOT has been applied to the description of patients with asthma, bronchitis, emphysema, diffuse interstitial lung disease and thoracic wall deformities, and to assess bronchial or therapeutic challenge [6, 29, 32, 36, 7478]. Multifrequency FOT using PRN imposes a more gentle forcing signal perturbation than IOS, and has not been noted to provoke bronchoconstriction. IOS differs from PRN by utilising brief pressure pulses of 30 40 ms duration. These pulses result in respiratory pressure responses that may be perceived as a slightly unpleasant respiratory sensation in some subjects. The brief pressure pulses provide convenient time-trend analyses and within-breath changes of Rrs and Xrs not available with PRN. IOS is most familiar to the current authors, and therefore, this and the following section relate current clinical work with IOS to previously published reports of both PRN and IOS. The clinical relevance of FOT may be assessed, as with any test of physiological function, in terms of its utility in diagnosis. Two general approaches are considered: rst, the use of FOT as part of an initial complete diagnostic evaluation, including spirometry, body plethysmography, and gas distribution and exchange measures; secondly, the use of FOT as a means of monitoring response to treatment can include both bronchial and therapeutic challenge. Summary information is presented here concerning the utility of FOT in assessing severity of lung disease, degree of airway reactivity, reversibility of airow obstruction, and stratication of breathing mechanics between central and peripheral airways. Current clinical relevance of FOT relates signicantly to the broad range of patients that may conveniently be evaluated. In contrast to standard PFT requiring maximal coordinated efforts, FOT requires only normal quiet breathing with the lips tightly closed to avoid airow leak, and the wearing of a nose-clip. For this reason, children can be easily studied, often as early as 3 yrs [8, 19, 9294]. Similarly, elderly subjects, those with severe airow obstruction or those with neuromuscular disease who nd maximal forced respiratory efforts difcult to perform are able to breathe normally for FOT testing [95 98]. Portability of commercial FOT instruments permits lung function testing at the bedside or, for occupational lung disease studies, at the place of work [18]. FOT places minimal performance demands upon the patient, often described as passive cooperation. However, the operator must take considerable care with the test procedure. Because of the freedom offered to patients by simply breathing normally, moment-to-moment changes in respiratory resistance may be anticipated. Accordingly, a minimum of three technically acceptable FOT tests of 2030 s duration or longer should be performed. The mouthpiece of the FOT instrument must be supported at a position to ensure maintenance of a neutral relaxed head and neck posture, avoiding body postures that might affect Zrs. Children should be seated in an appropriately-sized chair to comfortably support their legs and adults should avoid crossing their legs, which requires abdominal muscle contraction that may lead to end-expiratory lung volumes below relaxation volume. Patients should be comfortably relaxed to maintain a constant body
84
position without muscular effort. In contrast, rm contraction of the facial muscles is necessary to support airtight closure of the lips about the mouthpiece. It is also desirable that FOT testing be performed in a quiet examination room, sufciently far from others undergoing spirometry so as to be undisturbed by vigorous operator instructions for spirometry. The operator must review each FOT test immediately to ensure adequate recording time free of artefacts; a minimum of 20 s of consecutive artefact-free recording time is advisable. Lack of attention to these fundamental principles may result in highly variable FOT tests in an individual that are not clinically interpretable. This is not a troublesome issue in experienced clinical investigators laboratories but it is an important consideration in clinical PFT laboratories who have not previously used FOT. An important perceived concern about the clinical relevance of FOT is the availability of a normal database from which to judge results in a particular patient. Published reports of normative FOT data in children and adults are available, but the number and size of these studies represent a much smaller normal population than is available for spirometry [6, 32, 39, 74, 79, 80, 93]. Differences in techniques of FOT and more recently in mouthpiece design may allow some degree of uncertainty concerning ranges of normal expected values for both Rrs and Xrs, in both adults and children. Nevertheless, the similarity of FOT data using monofrequency sine-wave oscillations, PRN or pulseshaped multifrequency FOT over the past four decades supports the acceptance of clinically useful guidelines at this time. As expected, Rrs and Xrs are dependent on body size, and recent data suggest the possibility of racial/ethnic differences [99]. It is suggested that concern over precise denition of "normality versus abnormality" in an individual should not preclude clinical implementation of FOT at this time, because bronchial and therapeutic challenges are very helpful in assessing airway responsivity. If initial baseline Rrs/Xrs data do not clearly identify abnormality relative to existing normative data, retesting after b-agonist inhalation will immediately identify increased airway responsivity, and sequential studies over time provide similarly useful guidelines for clinical patient management.
Diagnostic evaluations
The most obvious relationship with other PFT procedures concerns the widespread use of spirometry. At the outset, it must be emphasised that spirometry measures maximal forced respiratory efforts, while FOT measures quiet breathing. Accordingly, it is not appropriate to demand that FOT and spirometric parameters be closely correlated as a mandatory requirement for FOT to be considered valid. For example, children with asthma most commonly manifest normal spirometry [100] with no spirometric response to inhaled b-agonist [42], while they may manifest abnormal baseline FOT parameters that are responsive to therapeutic challenge [101]. At the other end of the spectrum, patients with advanced chronic obstructive pulmonary disease (COPD) commonly manifest marked dynamic airway compression during spirometry with little spirometric response to pharmacological treatment, but may often manifest signicant FOT responses [45]. For these reasons, use of spirometry to dene severity of obstructive lung disease or receiver-operating characteristic of true sensitivity and specicity of oscillometry must no longer be considered the optimal standard. In some patients, spirometry cannot provide optimal clinical information. Patients with signicant neuromuscular disease are unable to provide the motive force needed for clear interpretation of spirometric results [97, 98]. Patients with lung allograft transplantation have obvious thoracic wall limitations that preclude truly maximal respiratory efforts until many months after surgery, during which time it is often not possible to detect adverse events, such as infection or acute allograft rejection by
85
H.J. SMITH ET AL.
spirometry alone. It is common for forced vital capacity (FVC) and FEV1 to increase over the rst 1826 weeks post-lung transplantation [102]. Despite this apparent improvement in spirometric parameters, peripheral airway disease, if it occurs, will lead to substantial worsening of FOT parameters [103] or of gas distribution [104]. Spirometric determination of responses to bronchial or therapeutic challenge may be limited by the necessary deep inspiration immediately prior to maximal expiration, allowing for distinctly different FOT responses [105108]. This disagreement between spirometric and FOT parameters during bronchial challenge may be readily documented by IOS during quiet breathing immediately before and after a deep inspiration, which commonly reveals immediate but transient bronchodilation in asthmatic subjects during bronchial challenge. Figure 5 illustrates a 70-s recording of tidal breathing and a deep inspiration with simultaneous display of calculated magnitude of impedance at 5 Hz (Zrs5) using impulse oscillometry. Rrs5 in this patient had increased markedly from 0.3 kPa?s?L-1 at baseline to 1.2 kPa?s?L-1 after cumulative inhalation of 0.25, 0.5 and 1.0 mg?mL-1 methacholine, when FEV1 had changed by only 260 mL from a baseline of 2.73 L. Zrs5 during normal breathing and following a deep inspiration with relaxed expiration shows that, at the onset of the IOS test, Zrs5 is much increased, while after an inspiration to maximal lung volume, there is a marked fall in Zrs5, which gradually "recovers" with time. Since deep inspiration to maximal lung volume must immediately precede the FEV1 measurement, patients like the one whose data are shown in gure 5 will manifest FEV1 responses to methacholine that are altered by the immediate decrease in airway smooth muscle tone following maximal inspiration [106, 107]. With these caveats it is suggested that FOT can provide useful supplementary information to spirometry that may not be tightly correlated with spirometric results. To put such FOT information into proper perspective, it is necessary to review commonly observed patterns of FOT results in lung disease (primarily airow obstruction), and in response to bronchial and therapeutic challenge.
3 Volume L Zrs5 kPasL-1 2 1 0 -1 2.0 1.6 1.2 0.8 0.0
0.4 0 10 20 30 40 Recording time s 50 60 70
Fig. 5. Tidal volume and magnitude of respiratory impedance at 5 Hz (Zrs5) plotted as a function of time during methacholine challenge. First 40 s are resting breathing. After 40 s, subject inspired to total lung capacity, followed by relaxation back to normal resting breathing. Note that Zrs5 increases markedly during each exhalation. After the deep inspiration, Zrs5 decreases transiently, with gradual return towards initial levels over the following 24 s. The moving average of Zrs5 is shown by the dash-dot line.
86
Oscillometry in relationship to other diagnostic pulmonary functional tests

An important body of work has related FOT to body plethysmography [29, 7678]. A group led by Van Noord have reported high correlations between Raw and FOT parameters in patients with obstructive lung disease and between absolute total lung capacity (TLC) and FOT parameters in patients with diffuse interstitial lung disease. In further studies comparing plethysmography, spirometry and FOT to assess reversibility of airow obstruction, Van Noords group reported the distinctly lower sensitivity of FOT than plethysmography [77]. The importance of this and other work by the Van Noord group is discussed further in the following sections. It is widely recognised that body plethysmographic resistance, Raw includes only the resistance of the extrathoracic and intrathoracic airways, while Rrs includes that of the chest wall and lung tissue in addition to airway resistance. Resistance of the chest wall has been reported [75], but there has been limited clinical interest in this parameter because of the technical difculty of the measurements. Another difference between Raw and Rrs relates to the status of the glottic aperature: it is commonly assumed that the glottis is maximally open during panting, but Jackson et al. [109] have shown that this occurs only in totally unrestricted panting. During voluntary attempts to control panting frequency and tidal volume, there is signicant adduction of the vocal cords. Similarly, during quiet breathing, normal subjects commonly manifest a small, but variable, degree of vocal cord adduction during expiration. In patients with obstructive lung disease, this phasic expiratory adduction, visualised during the course of bronchscopy, does not appear to be systematically different from that observed in normal subjects. Thus, it is to be expected that average Rrs will differ systematically from panting plethysmographic Raw. It is also well established that Raw is more prominently inuenced by large airway than by small airway resistance. Thus, Smith and Dubois [110] reported a comparable increase in deadspace when compared to the decrease in Raw in response to scopolamine in normal subjects. In addition, Hensley et al. [111] reported similar changes in Raw and deadspace after inhaled atropine. These results are consistent with the idea that Raw is primarily inuenced by large airways. In contrast, Rrs is importantly inuenced by small airway resistance, and, accordingly, it may be expected that FOT responses to interventions that improve peripheral airway obstruction will be more prominent than Raw responses. Because of the prominent effect of peripheral airway obstruction on FOT measurements, it may be expected that FOT indices of peripheral airway obstruction might correlate more closely with indices of gas distribution ("Closing volume" [104]) and areas of lung hyperination manifested by computerised tomography [112] than with plethysmographic or spirometric indices, although there are no published comparisons at this time.
Oscillometry as a clinical monitor of response to treatment

By way of summarising the clinical relevance of FOT, it is worth considering the special value of FOT as a means of monitoring response to interventions. FOT has been reported to show greater sensitivity to inhaled corticosteroid or to b-agonist inhalation [8, 113115] than spirometry. Both inhaled corticosteroids and b-agonists improve small airways function, and FOT responses manifest prominent changes in indices of peripheral airway obstruction. In contrast, spirometric sensitivity to small airways function is less prominent. Accordingly, it is expected that FOT might provide useful indices of peripheral airway change in response to therapeutic interventions. Such use of FOT provides a clinically valuable monitoring tool to follow therapeutic changes in small
87
H.J. SMITH ET AL.
airways function over time. This use of FOT for therapeutic monitoring is not dependent on the use of FOT as an initial diagnostic evaluation. Finally, as a matter of practical convenience, FOT is more readily utilised in the clinical pulmonary function laboratory than body plethysmography. This issue is relevant to recent interest in the therapeutic value of anticholinergics in patients with COPD. As noted above, anticholinergics result primarily in large airway bronchodilation, and changes in Raw and deadspace are considerable [110, 111]. Thus, effective airway cholinergic blockade decreases large airway bronchomotor tone and increases deadspace, with relatively little effect on small peripheral airways disease. Whereas body plethysmography may be considered a useful technique to monitor such treatment effects, it is substantially less convenient to use routinely than FOT.
Oscillometry in the clinical pulmonary laboratory

Clinical interpretations of FOT responses in patients have often been related directly to the application of a particular mechanical or electrical model of the respiratory system. Van Noord et al. [78] discussed their results in diffuse interstitial lung disease with respect to an electrical analogue of the respiratory system. They further conrmed earlier work (vide infra) that ascribed negative frequency dependence of resistance to peripheral airway obstruction [116, 117]. Engineering models of the respiratory system have provided predictions of FOT characteristics in normal human subjects, and changes in FOT parameters in lung disease. However, the fact that many of these predictions have been observed empirically does not constitute proof of validity of one or other engineering models. Rather, it provides evidence that under the particular conditions of the FOT measurements undertaken, the empirical results show patterns that would be intuitively expected in normal subjects and those with lung disease. Over the past 3 decades, a body of empirical evidence has accumulated that relates FOT results to particular lung diseases, indeed establishing patterns that are characteristic of lung disease. The following sections draw heavily upon this clinical research and codify FOT results with respect to obstructive lung disease, with very limited data in diffuse interstitial lung disease. These FOT data are not intended as validations of engineering models, but instead, to illustrate commonly observed patterns of FOT characteristics associated with lung disease.
Obstructive lung disease

The relationships of FOT to spirometry noted above have a common theme. Spirometry does not provide a clear indication of peripheral airway obstruction, despite the general information contained within the shape of the owvolume curve, and values of mid-ow rates (forced expiratory ow between 25 and 75% of the forced vital capacity). Thus, the most striking characteristic of FOT in relation to spirometry is the relatively greater sensitivity of FOT to peripheral airway disease [2, 18, 25, 29, 32, 42, 45, 52, 68, 82].
Peripheral airway disease. The most well-known FOT result empirically observed in
peripheral airway disease in adults is frequency dependence of resistance (fdr). Grimby et al. [116], using multiple replicates of monofrequency FOT, were the rst to demonstrate the pattern of frequency dependence, wherein calculated Rrs was greater at 3 Hz than at 5, 7 or 9 Hz in patients with chronic airow obstruction (CAO). As calculated Rrs decreased as oscillation frequency increased, patients with CAO might manifest nearly normal
88
values of Rrs at sufciently high frequencies. For this reason, Grimby et al. [116] focused on low (3 Hz) monofrequency Rrs to avoid masking differences between patients with airow obstruction and normal subjects [118]. Many subsequent reports [18, 32, 36, 74, 82, 117, 119] conrmed that subjects with early peripheral airways disease, including smokers, certain industrial workers and normal subjects after histamine infusion, manifested frequency dependence, even with normal values of low-frequency Rrs in smokers. Importantly, the abnormal frequency dependence of resistance occurred in the presence of normal spirometry in those subjects with early peripheral airways disease [18, 32, 117]. This body of clinical research is largely empirical, although it had been shown on autopsy many years earlier that cigarette smokers who died early in life had manifested peripheral airway inammation on autopsy [120]. Similarly, there is now ample evidence of peripheral airway inammation in patients with asthma, and, as will be illustrated below, frequency dependence of resistance occurs prominently in asthma. The sensitivity of frequency dependence to peripheral airway disease is the rst discriminant between methods of FOT in general: while monofrequency FOT is convenient to dissect within-breath patterns of change in Rrs [61, 68] or changes in Rrs during sleep-disordered breathing or in patients on mechanical ventilators [2], multifrequency FOT is most convenient to document frequency dependence of resistance in practical use in the clinical pulmonary function laboratory. Monofrequency FOT may be used at two or more single frequencies; however, multifrequency FOT uses different oscillation frequencies applied within a single burst to dissect patterns consistent with peripheral rather than central airway obstruction. This dissection is based upon established observations that pressure oscillations at frequencies w15 Hz are severely damped out before reaching peripheral airways, while those at frequencies v1015 Hz penetrate much further to the lung periphery [25, 121]. The transition between "large central" airways and "small peripheral" airways is neither precisely xed anatomically nor precisely dened in terms of airway lumen diameter. The illustrations in gures 6 and 7 reect common patterns observed in children with asthma and in adults with COPD. Figure 6 shows representative IOS tests pre- and post-salbutamol in a 6-yr-old patient
1.0 0.8 0.6 0.4 0.2 0.0
Resistance Rrs kPasL-1
10
20 15 Frequency Hz
25
30
35
Fig. 6. Conventional plots of respiratory resistance (Rrs(f)), as a function of oscillation frequency, in a 6-yr-old child with asthma. Note that frequency axis is shown between zero and 35 Hz, while data are plotted between 335 Hz. A vertical dotted line is shown at 5 Hz, the lower limit at which most impulse oscillometry system data are reported. : data prior to nebulised b-agonist bronchodilator; ----: data after bronchodilator.
89
H.J. SMITH ET AL.
1.0 0.8 0.6 0.4 0.2 0.0
10
20 15 Frequency Hz
25
30
35
Fig. 7. Respiratory resistance (Rrs(f)) in an adult patient with chronic obstructive pre- and post-nebulised bagonist bronchodilator. : pre-bronchodilator; ----: post-bronchodilator. Note that Rrs(f) is unchanged after bronchodilator at frequencies w12 Hz.
with mild asthma. Rrs is plotted as a function of oscillation frequency (Rrs(f)-tracing). Note that prior to b-agonist, Rrs5 is 0.93 kPa?s?L-1. Rrs falls steeply with increasing oscillation frequency to a minimum at 18 Hz, after which it increases with further increases in oscillation frequency. While Clement et al. [39] have shown that normal children manifest a mild degree of frequency dependence, the very large fall in Rrs between 5 and 15 Hz in gure 6 is consistent with abnormal peripheral airways function in a 6 yr old. This is conrmed by administration of nebulised b-agonist, after which Rrs5 decreased to 0.59 kPa?s?L-1 (37% change). Note also that the fall of Rrs between 515 Hz post b-agonist is much less than pre b-agonist. Baseline and post b-agonist IOS data in this child may be compared with data in normal children of this age, who manifested v15% fall in Rrs from 515 Hz [122]. The response to b-agonist in gure 6 may also be compared with responses of normal nonatopic children who manifested an average change in Rrs5 after salbutamol of 19% [122]. Finally, it can be seen in gure 6 that Rrs(f) at frequencies w20 Hz decreased substantially after b-agonist. Adult patients with asthma show similar ndings to those in gure 6. Rrs may be nearly independent of oscillation frequency in adult asthmatics after beta agonist. Figure 7 illustrates an adult patient with COPD, pre- and post-b-agonist. In contrast to asthma, Rrs(f) decreases continuously with oscillation frequency between 525 Hz at baseline, and after b-agonist there is a decrease in Rrs(f) only at low frequencies,v12 Hz. The ndings in gures 6 and 7 are consistent with bronchodilation produced by bagonist in both large and small airways in the asthmatic subjects, but only in small airways in the patient with COPD. In some patients with COPD who also have reactive airways, b-agonist results in bronchodilation of large airways as well.
Central airway obstruction. Because of the frequency-dependent distribution of

oscillatory pressures within the airway tree, FOT provides separate, although not entirely independent, indices of large and small airway responses. Thus, changes of resistance in the larger airways are manifest by FOT as uniform changes in Rrs at all oscillation frequencies, both low and high. An increase in resistance of the large airways was reported in a recent study of rescue workers at the World Trade Center site in New York [18]. Rescue workers with no history of cigarette smoking exposed to large-particle air
90
pollution at the World Trade Center site manifested, at baseline, uniformly increased Rrs at all oscillatory frequencies studied between 535 Hz, as shown in gure 8, pre- and postnebulised b-agonist. Ironworkers with a history of cigarette smoking showed greater baseline increases in low-frequency Rrs5, and, accordingly, a frequency dependence of resistance that is characteristic of peripheral airway obstruction [7, 18, 29, 36, 82, 116, 117]. Responses to nebulised b-agonist showed uniform decreases in Rrs across all frequencies in nonsmokers, while smokers manifested larger decreases in Rrs5 than in Rrs20 [18].
Nonresistive components of forced oscillation technique. A second discriminant

between mono- and multifrequency FOT concerns reactive components of applied pressure oscillations, reactance Xrs(f)-tracings, which appear prominently in multifrequency FOT. Reactance can be assessed by computer-assisted analyses utilising FFT; however, monofrequency FOT has not been widely adapted to conveniently calculate Xrs. Just as with the interpretation of Rrs, oscillation frequency provides a means of examining different regions of the airway tree using Xrs (vide infra): at low oscillation frequencies, elastic elements in peripheral airways are the dominant reactant to applied pressure, and reect small airway mechanical properties. In airow obstruction, small airways are functionally obstructed, due to peripheral airway inammation in both asthma and COPD. This results in portions of the distal lung periphery that are "in the shadow" of effective obstruction of small airways. This pathological process leads to an increase in the magnitude of Xrs at low frequencies. At high frequencies, accelerative forces are the predominant reactant to applied pressures and occur virtually exclusively in large airways where they are related to inertial properties. It should be noted that Van Noord et al. [78] reported qualitatively similar changes in Xrs(f) tracings in patients with diffuse interstitial lung disease and obstructive lung disease. Accordingly, changes in lowfrequency Xrs are not specic to obstructive lung disease, but rather reect peripheral airway disease. Mechanical interpretations of changes in low-frequency Xrs in obstructive and restrictive lung disease are considered in detail in the methodology section.
1.0 0.8 0.6 0.4 0.2 0.0
10
20 15 Frequency Hz
25
30
35
Fig. 8. Respiratory resistance (Rrs(f)), plotted as a function of oscillatory frequency pre- () and post (----) -bronchodilator in an ironworker exposed to large-particle air pollution at the World Trade Center site. Note increased Rrs at all frequencies, with no signicant frequency dependence of resistance at baseline. After nebulised b-agonist, Rrs(f) decreases in a parallel manner relative to baseline pre-bronchodilator.
91
H.J. SMITH ET AL.
Figure 9 illustrates IOS Xrs(f) tracings in the asthmatic child whose Rrs(f) tracings preand post-nebulised b-agonist are shown in gure 6. Figure 9 shows that after b-agonist, Xrs at 5 Hz (Xrs5) changes from -0.31 to -0.26 kPa?s?L-1 and the frequency at which Xrs is zero (resonant frequency = fres, vide infra) changes from 18 to 17 Hz (6%). More strikingly, the overall curvature of the Xrs(f) tracing changes from concave to the zero-X axis to being convex to the zero-X axis. This change in curvature is consistent with Xrs(f) tracings described by Clement et al. [36], who reported that patients with airow obstruction manifested a loss of the downward concavity of Xrs(f) tracings that is commonly seen in normal subjects. b-agonist produced only small changes in Xrs5 and fres in gure 9; however, the change in the overall Xrs(f) tracing curvature at frequencies below fres was dramatic. Previous investigations have emphasised that low-frequency Xrs and Rrs are most sensitive to changes in peripheral airway function, and Xrs5 has been used as a primary efcacy variable [113115, 122, 123]. However, in small children, respiratory frequency is commonlyw2030 breaths?min-1, and, accordingly, the higher harmonics of fundamental respiratory frequency may encroach on the lowest FOT frequencies analysed [41, 68]. As a result, Xrs5 manifests relatively greater measurement noise. Some IOS studies in children have reported failure of Xrs5 to manifest signicant changes after inhaled corticosteroids or b-agonists [42, 115, 124]. Marotta et al. [42] showed that the Xrs10 response to b-agonist, but not the Xrs5 response, manifested a signicant difference between asthmatic and nonasthmatic atopic children. This was associated with less variability in Xrs10 responses compared with Xrs5. The absolute value of Xrs (|Xrs|) changes differently as a function of oscillation frequency below and above fres. At low frequencies of oscillation, below fres, |Xrs| decreases as oscillation frequency increases up to fres. At fres, |Xrs| is zero. As oscillation frequency increases above fres, |Xrs| increases with further increases in oscillation frequency. Because of this difference in the relationship between magnitude of Xrs and oscillation frequency below and above fres, a quantitative index of Xrs magnitude at frequencies below fres was developed by integrating all negative values of Xrs [18, 45, 52]. This index,
0.5 0.4 Reactance Xrs kPasL-1 0.3 0.2 0.1 0 -0.1 -0.2 -0.3 -0.4 -0.5 0 5 10 AX pre 15 20 Frequency Hz 25 30 35 AX post
Fig. 9. Respiratory reactance (Xrs(f)), plotted as a function of oscillatory frequency pre- () and post- (----) bronchodilator. Same subject as gure 6. The integrated low-frequency reactance area (AX), is shown by vertical hatching pre-bronchodilator, and by diagonal hatching post-bronchodilator. This area is reduced by y50% from pre- to post-bronchodilator, associated with marked change in curvature of the Xrs(f) tracing. In comparison, there are small changes in Xrs at 5 Hz and resonant frequency from pre- to post-bronchodilator.
92
designated AX, provides an integrative function to include changes in the magnitude of low-frequency Xrs, changes in fres and changes in curvature of the Xrs(f) tracing. AX includes Xrs magnitudes at 5 Hz and slightly higher oscillation frequencies which manifest improved signal-to-noise ratio, as noted above for atopic asthmatic children [42]. It is represented graphically in gure 9 as the area under the zero Xrs axis above the Xrs(f) tracing. As discussed below, this integrative index provides a single quantity that reects changes in the degree of peripheral airway obstruction and correlates closely with frequency dependence of resistance. In gure 9, AX decreases by 50% after b-agonist, closely comparable to the decrease in frequency dependence of resistance measured between 5 and 15 Hz in the same child shown in gure 6, when R5-R15 decreased from 0.36 kPa?s?L-1 to 0.16 kPa?s?L-1 after b-agonist. The perspective presented here of AX in relation to peripheral airway function, results directly from the details presented in the methodology section, namely that lowfrequency Xrs essentially expresses the ability of the respiratory tract to store capacitive energy, which is primarily resident in the lung periphery. In contrast, at frequencies above fres, I, which is primarily resident in proximal conducting airways, contributes signicantly to the dissipation of externally applied pressures [6, 46]. Thus, oscillation frequencies above fres reect mechanical properties of more proximal conducting airways. Because of these differences in mechanical properties reected by low- and highfrequency oscillation, calculation of the arithmetic mean Xrs is not likely to be optimally useful to assess pulmonary mechanical responses. Thus, Van Noord et al. [29] reported that the mean value of Xrs change was less sensitive than FEV1 in assessing the effect of histamine. However, their graphic mean Xrsfrequency data reveal changes in the estimate of AX from y0.15 at baseline, to 1.7 or 2.5 kPa?L-1 when mean decrease in specic airway conductance (sGaw) was 40% or 15% in FEV1. These increases in AX of 1,0001,500% are comparable to those measured using IOS in the current authors laboratory during methacholine challenge: the patient shown in gure 6 manifested an increase in AX from 0.34 at baseline to 7.5 kPa?L-1 (w2,000% increase) after cumulative exposure to methacholine up to 1.0 mg?mL-1. Furthermore, the study of Van Noord et al. [77] during assessment of reversibility of airow obstruction by FOT, body plethysmography and spirometry reported that changes in mean Xrs did not contribute signicantly to discriminant function beyond spirometry, Raw and Rrs at 6 Hz. In contrast, estimated AX, approximated from their graphic mean Xrsfrequency data, showed a 50% reduction after salbutamol, from y3.1 to 1.5 kPa?L-1. Both the baseline AX in patients with airow obstruction and the AX decreases in response to b-agonist are comparable to those recently reported using IOS [101]. The empirical observations discussed above do not prove that any particular engineering model is a true representation of the lung, especially the diseased lung as emphasised by Van Noord et al. [78]. Nevertheless, FOT results predicted by models may correlate usefully with independent clinical physiological and pathophysiological evidence. Quite apart from engineering models, an intuitive understanding of Xrs may be appreciated from the physical principles elucidated above and amplied in the foregoing discussion of methodology. The applicability of high frequencies to large central airways and low frequencies to peripheral airways is not a consequence of any particular engineering model, but is observed empirically both in Rrs and Xrs.
Clinical interpretation of forced oscillation technique

As also noted in the methodology discussion of FOT, abnormalities of Xrs are not specic to obstructive lung disease, because these same patterns have been reported in lung brosis [78]. In clinical diagnostic lung function testing including FOT, spirometry,
93
H.J. SMITH ET AL.
gas diffusion and body plethysmography, the problem is not usually distinguishing between obstructive and restrictive disease. The more important issue is the relative severity of pathophysiological abnormality. Furthermore, there is evidence to suggest that the early pathological lesion in lung brosis is inammation of the small airways [125]. Thus, changes in Xrs magnitude in lung brosis and in peripheral airow obstruction may both reect peripheral airway inammation. If these nonspecic Xrs abnormalities in lung brosis represent small airway inammation, they may respond to anti-inammatory treatment, analogous to the way asthmatic peripheral airway inammation responds to corticosteroids. Such responses are more likely to be found in FOT parameters than in spirometry. Clinical interpretation may then be considered in the setting of response to treatment interventions. Clinical interpretation of FOT can be related both to effects of airway smooth muscle tone and airway inammation. Below, effects of anticholinergic, b-agonist or corticosteroid medications are represented, because these agents are most commonly utilised clinically. Commonly observed changes in FOT measures in patients with obstructive lung disease are illustrated. Rrs is considered rst.
Rrs effects. While inammation is a cellular process, it has mechanical consequences. These consequences may be considered in relation to bronchoconstriction, dened as increased tone of airway smooth muscles, and the common perception of bronchodilation dened as a decrease in smooth muscle tone. When airway smooth muscle tone increases, Rrs increases because of decreased airway lumen. Airway lumen is also decreased with inammation or oedema in the walls of the airways. Therefore, Rrs increases as a result of inammation and oedema. Peripheral airways have much smaller lumina than central (large) airways, and inammation/ oedema in the walls of peripheral airways can reasonably be expected to have a proportionately larger effect on lumen size than inammation/oedema in larger airways. In the discussion that follows, "low-frequency Rrs" will be denoted as Rrs at frequencies v15 Hz and "high-frequency Rrs" as Rrs at frequencies w20 Hz, with the latter term synonymous with large airway resistance. If an intervention, such as inhaled anticholinergic, achieves bronchodilation with no effect on inammation, it may be expected that large airway lumen will increase, with little effect on peripheral airway lumen. In this event, large airway Rrs will decrease. Lowfrequency Rrs will decrease to a similar degree and little or no change in frequency dependence occurs. This is illustrated in a 55-yr-old patient with COPD in gure 10. If an intervention such as inhaled b-agonist achieves bronchodilation with little or no effect on inammation, it may be expected that peripheral airway lumina will increase, in addition to any release of bronchoconstriction in large airways. In this case, lowfrequency Rrs will decrease out of proportion to high-frequency Rrs. In asthmatic patients, both high-frequency and low-frequency Rrs may decrease, with relatively greater decrease in low-frequency Rrs and associated decrease in frequency dependence of resistance as illustrated in gure 6. In patients with COPD, a decrease in low-frequency Rrs after b-agonist with little or no change in high-frequency Rrs is commonly observed, as illustrated in gure 7. In COPD patients with lung hyperination, little or no decrease in Rrs may occur after b-agonist inhalation, particularly if there is an associated fall in resting end-expiratory lung volume. If Rrs remains the same after b-agonist while end-expiratory lung volume decreases, this represents "functional bronchodilation", because the same resistance pertains at lower operating lung volumes. Accordingly, failure of Rrs to decrease in patients with COPD need not be considered as "no response" to b-agonist bronchodilator. If an intervention such as inhaled corticosteroids achieves a decrease in inammation
94
2.0 1.6 1.2 0.8 0.4 0.0
10
20 15 Frequency Hz
25
30
35
Fig. 10. Respiratory resistance (Rrs), plotted as a function of oscillation frequency in a 55-yr-old male with chronic obstructive pulmonary disease pre- () and post- (----) anticholinergic bronchodilator. Note that Rrs at 5 Hz is markedly elevated with marked frequency dependence of Rrs. After inhaled anticholinergic bronchodilator, there is a signicant decrease in Rrs that is nearly identical at all frequencies, indicating a decrease in proximal airway resistance, with little or no effect on peripheral airways
with no effect on airway smooth muscle tone, FOT responses might be expected to reect a relatively greater impact due to decrease in peripheral airway inammation with resultant increase in peripheral airway lumina. Such an effect results in a signicant "dilation" of peripheral airways due to decreased inammatory encroachment on peripheral airway lumina. Thus, a decrease in peripheral airway resistance can be expected, manifest as a greater decrease in low-frequency than in high-frequency Rrs, and concomitant decrease in frequency dependence. In asthmatic patients, a decrease in large airway Rrs (high-frequency Rrs) may also occur. In patients with COPD, there may be a decrease in low-frequency Rrs; however, little or no decrease in Rrs may be manifest, especially if lung hyperination is present.
Xrs effects. How then does decreased inammation manifest itself in patients with COPD? As noted in the preceding section, describing nonresistive components of FOT, the magnitude of low-frequency Xrs is increased in COPD due to functional peripheral airway obstruction, with resultant contraction of surface area of the lung periphery exposed to low-frequency oscillations. Indeed, low-frequency Xrs is more sensitive to peripheral airway obstruction in COPD/emphysema than Rrs. In the presence of peripheral airway obstruction in patients with COPD, relatively small increments to airways resistance may occur because of the large cumulative cross-sectional diameter of all airways in the lower generations of airways, as manifest in the trumpet model of Weibel [126]. Accordingly, body plethysmographic measurement of airway resistance may be nearly normal, and only by measuring absolute thoracic gas volume is the abnormality manifest. If peripheral airway inammation is decreased by administration of inhaled corticosteroids, peripheral airway lumina increase and the patency of small airways expands in the direction of the lung periphery. As a result, a portion of the lung periphery comes "out of the shadow" of small airway obstruction and a larger surface area is presented to the low-frequency oscillations. This acts to decrease the magnitude of lowfrequency Xrs. This is illustrated in gure 11a showing IOS tracings in a COPD patient at
95
H.J. SMITH ET AL.
a) 0.5 0.4 Reactance Xrs kPasL-1 0.3 0.2 0.1 0.0 -0.1 -0.2 -0.3 -0.4 -0.5 b) 1.0 0.8 0.6 0.4 0.2 0.0 AX pre AX post
10
15 20 Frequency Hz
25
30
35
Fig. 11. a) Respiratory reactance (Xrs), plotted as a function of oscillatory frequency in a 73-yr-old male with chronic obstructive pulmonary disease before and after 4 weeks inhaled corticosteroid (ICS) therapy. : preICS; ----: post-ICS. Integrated low-frequency reactance area (AX) is vertically hatched pre-ICS and diagonally hatched post-ICS. AX is decreased by y50% after 4 weeks of therapy, resonant frequency by 10% and Xrs at 5 Hz decreased by 0.12 kPa?s?L-1. b) Respiratory resistance (Rrs) plotted as a function of oscillation frequency in the same patient. : pre-ICS; ----: post-ICS. Note that Rrs at 2022 Hz is relatively unchanged, while Rrs at 5 and 10 Hz are substantially decreased. High-frequency resistance (Rrs at 2535 Hz) is decreased after ICS therapy. See text for discussion.
baseline () and after 4 weeks of inhaled corticosteroids treatment (-----). AX decreased from 3.0 to 1.3 kPa?L-1 after inhaled corticosteroid treatment. In this patient, Rrs and frequency dependence of the Rrs(f) tracing show signicant decreases, as illustrated in gure 11b; Rrs5 improved from 0.68 to 0.48 kPa?s?L-1, Rrs15 from 0.41 to 0.35 kPa?s?L-1. Furthermore, this patient with COPD manifests somewhat "responding" large airways, as his high-frequency Rrs (at 3035 Hz) also decreased by y20%. Figure 12 shows tracings in a COPD patient when stable at baseline and during the onset of an exacerbation. Figure 12a shows baseline reactance area AX = 1.2 kPa?L-1, with increases to 1.6, 2.4 and 4.3 kPa?L-1 over a duration of 9 days of exacerbation. Rrs5 increased signicantly, but less dramatically from 0.51 at baseline to 0.54, 0.59 and 0.65 kPa?s?L-1 during exacerbation, shown in gure 12b. Frequency dependence, calculated as the fall in Rrs from 5 to 15 Hz, was 0.15 kPa?s?L-1 when stable at baseline, and increased to 0.2, 0.23 and 0.31 kPa?s?L-1 during exacerbation.
96
a) 0.5 0.4 Reactance Xrs kPasL-1 b) Resistance Rrs kPasL-1 0.3 0.2 0.1 0.0 -0.1 -0.2 -0.3 -0.4 -0.5 1.0 0.8 0.6 0.4 0.2 0.0
10
15 20 Frequency Hz
25
30
35
Fig. 12. a) Respiratory reactance (Xrs(f)) in a 73-yr-old male chronic obstructive pulmonary disease patient when stable at baseline and during onset of exacerbation over a duration of 9 days. Note progressive increase in the reactance area (AX) from trial 1 to trial 4. b) Respiratory resistance (Rrs(f)) in the same patient. Note progressive increase in Rrs at 5 Hz (Rrs5) from trial 1 to trial 4. Changes in Rrs5 were relatively smaller than changes in AX over a duration of 8 days. : trial 1, baseline; -----: trial 2, 7 days after trial 1; - : trial 3, 8 days after trial 1; -- : trial 4, 9 days after trial 1.
The close correlation between changes in AX and changes in frequency dependence of resistance in individual patients shown in gures 612 are conrmed across individuals in the occupational study reported by Skloot et al. [18], who found a correlation of 0.92 between frequency dependence of resistance and AX across a sample of ironworkers at the World Trade Center site, with variable exposure to cigarette smoking and largeparticle air pollution, and resultant variability in both large and small airway obstruction. The close correlation between frequency dependence of resistance and AX is consistent with both indices reecting small airway function.
Coherence. Coherence, rst introduced by Michaelson et al. [5], is dened as the autoand cross-correlations of phase and amplitude of oscillatory pressure and ow components. It reects, in an "engineering sense", the linearity of the respiratory system and, in a "biological sense", the variability of the respiratory system from time to time within the sample of data. Marked temporal variability of the respiratory system within a breath occurs commonly in COPD, where even during quiet breathing, dynamic
97
H.J. SMITH ET AL.
compression of intrathoracic airways may occur. As a result, Rrs and Xrs may increase substantially during expiration. Figure 13a illustrates a patient with severe COPD, with volume and Zrs5 as functions of time. Figure 13b shows coherence plotted as a function of oscillation frequency, with separate tracings for average combined inspiration/expiration as well as for inspirationonly and expiration-only. Figure 13a shows marked changes in Zrs5 with respiratory phase, similar to those shown by Marchal and Loos [68]. There is a clear decrease in Zrs5 at the onset of inspiration, keeping minimal values until end-inspiration. A marked abrupt rise in Zrs5 occurs at the onset of expiration with elevated values including end-expiration. In gure 13b, the value of averaged coherence at 5 Hz, 0.4, is distinctly lower than at frequencies i10 Hz.
a) 1.5 Volume L Zrs5 kPasL-1 1.0 0.5 0.0 -0.5 3 2 1 0 b) 1.0 0.9 0.8 Coherence g2 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0 0 10 20 Time s 30 40
10
15 20 Frequency Hz
25
30
35
Fig. 13. a) Volume and magnitude of respiratory impedance at 5 Hz (Zrs5) plotted as a function of time during a 40-s impulse oscillometry system (IOS) test in a 53-yr-old patient with severe chronic obstructive pulmonary disease. Note marked increase in Zrs5 during the expiratory phase of every tidal volume. b) Coherence of all IOS data (global average () and separated inspiratory ( - ) and expiratory (-----) respiratory phases) plotted as a function of oscillation frequency during the 40-s test. Note low coherence for average data (v0.5) at 5 Hz with prominent increase to 0.7 at 10 Hz. See text for discussion.
98
Numerical calculations of coherence during the separate respiratory phases reveal that inspiratory-only and expiratory-only coherences are systematically greater than the combined coherence over the entire spectrum of oscillatory frequencies. The tracings in gure 13 are consistent with more uniformity of respiratory mechanics within the separate inspiratory and expiratory phases than for the combined total breath average. At 10 Hz, the coherence averaged across both respiratory phases is 0.7, while that for separate inspiratory and expiratory data are both 0.9, reecting differences in respiratory mechanical parameters that pertain to the separate respiratory phases. Despite the very low coherence for combined inspiration/expiration, three consecutive 40-s impulse oscillometry system recordings obtained within 5 min manifested an average respiratory resistance at 5 Hz of 1.08, 1.0 and 1.02 kPa?s?L-1 and reactance area of 5.9, 6.1 and 6.2 kPa?L-1. Thus, the within-phase uniformity of mechanical parameters reected by separate inspiratory and expiratory coherences is borne out by standard deviations of v5% for triplicate measures.
Summary
The aim of this chapter has been to describe the unique and clinically relevant information that forced oscillation technique (FOT) provides. This may be derived without mathematical mastery of technological principles of the equipment and/or of numerical models. It is emphasised that recognition of the change in respiratory mechanical parameters as a function of oscillation frequency is necessary to appreciate the outstanding value of FOT in its ability to assess peripheral airway function. This has been one of the major challenges in respiratory diagnostics up to the present time. The short duration of the FOT test, 2030 s, makes it particularly useful as part of a diagnostic programme of lung function testing; it is not suggested that FOT be used in lieu of conventional pulmonary function testing, but rather in addition. FOT measures resting breathing while spirometry assesses maximal respiratory performance of the patient. The special value of FOT in terms of short-term response to bronchial and therapeutic challenge has been emphasised as well as its value in monitoring long-term trend responses to therapy. The simplicity of FOT measurements and its minimal requirements on subject cooperation are in rather sharp contrast to its current limited clinical acceptance. Two primary reasons for the present limited application of FOT include the need for viewing respiratory mechanical parameters over a range of frequencies and the resultant central-peripheral specicity of oscillatory parameters, with specic emphasis on the reection of peripheral airway function by low-frequency reactance. Indeed, lack of awareness of this ability of FOT to assess peripheral airway function has turned physicians to the use of multiple replicates of high-resolution computed tomography lung scans to assess small airway function. Other reasons for limited use of FOT currently may include the greater variability of FOT measures compared with spirometry. Despite such variability, use of at least three replicate FOT measures combined with therapeutic challenge can provide sensitive evaluation of small airway function. The freedom allowed to the subject to breathe "naturally" imposes increased demands for vigilance on the operator, who must maintain a quiet environment for forced oscillation technique testing. Operators must also reassure subjects that their relaxation is needed, except for the facial musculature ensuring tight lip closure on the mouthpiece. Posture must be supported to maintain subject comfort and the
99
H.J. SMITH ET AL.
instrument mouthpiece must be brought to the subject to avoid stretching of the neck. Finally, the availability of results from a brief test must not lead the operator to accept a single measurement, but rather, the usual clinical testing procedure of at least three replicate measures is required. Keywords: Forced oscillation technique, impulse oscillation technique, reactance, resistance, respiratory impedance.
References
1. 2. 3. DuBois AB, Brody AW, Lewis DH, Burgess BF. Oscillation mechanics of lungs and chest in man. J Appl Physiol 1956; 8: 587594. Navajas D, Farre R. Oscillation mechanics. Eur Respir Mon 1999; 4 (12): 112140. Bedenice D, Mazan M, Reinhold P, Harrington K, Hoffman A. Comparison of two different oscillatory techniques for the non-invasive measurement of respiratory system resistance (Rrs) in the horse. Eur Respir J 2003; 22: Suppl. 45, 688. Lorino A, Zerah F, Mariette C, Harf A, Lorino H. Respiratory resistive impedance in obstructive patients: linear regression analysis vs viscoelastic modelling. Eur Respir J 1997; 10: 150155. Michaelson E, Grassman E, Peters W. Pulmonary mechanics by spectral analysis of forced random noise. J Clin Invest 1975; 56: 12101230. Landser F, Nagels J, Demedts M, Billiet L, Van De Woestijne K. A new method to determine ` frequency characteristics of the respiratory system. J Appl Physiol 1976; 41: 101106. Pride N. Forced oscillation techniques for measuring mechanical properties of the respiratory system. Thorax 1992; 47: 317320. Delacourt C, Lorino H, Herve-Guillot M, Reinert P, Harf A, Housset B. Use of the forced oscillation technique to assess airway obstruction and reversibility in children. Am J Respir Crit Care Med 2000; 161: 730736. Cauberghs M, Van de Woestijne K. Effect of upper airway shunt and series properties on respiratory impedance measurements. J Appl Physiol 1989; 66: 22742279. Daroczy B, Hantos Z. Generation of optimum pseudorandom signals for respiratory impedance measurements. Int J Biomed Comput 1990; 25: 2131. Solymar L, Landser F, Duiverman E. Measurement of resistance with the forced oscillation ` technique. Eur Respir J 1989; 2: Suppl. 22, 150153s. MacLeod D, Birch M. Respiratory input impedance measurement: forced oscillation methods. Med Biol Eng Comput 2001; 39: 505516. Vogel J, Smidt U. Impulse Oscillometry Analysis of Lung Mechanics in General Practice and the Clinic, Epidemiological and Experimental Research. Frankfurt, Pmi Verlagsgruppe GmbH, 1994. Peslin R, Jardin P, Duvivier C, Begin P. In-phase rejection requirements for measuring respiratory input impedance. J Appl Physiol 1984; 56: 804809. Van de Woestijne K, Desager K, Duiverman E, Marchal F. Recommendations for measurement of respiratory input impedance by means of the forced oscillation method. Eur Respir Rev 1994; 4: 235237. Lorino H, Mariette C, Karouia M, Lorino A. Inuence of signal processing on estimation of respiratory impedance. J Appl Physiol 1993; 74: 215223. Oostveen E, MacLeod D, Lorino H, et al. The forced oscillation technique in clinical practice: methodology, recommendations and future developments. ERS Task Force. Eur Respir J 2003; 22: 10261041. Skloot G, Goldman M, Fischler D, et al. Respiratory symptoms and physiologic assessment of ironworkers at the World Trade Center disaster site. Chest 2004; 125: 12481255.
4. 5. 6. 7. 8.
9. 10. 11. 12. 13.
14. 15.
16. 17.
18.
100
19. 20.
21.
22. 23. 24. 25.
26. 27.
28. 29.
30.
31.
32.
33. 34.
35. 36. 37.
38.
Bisgaard H, Klug B. Lung function measurement in awake young children. Eur Respir J 1995; 8: 20672075. Reinhold P, Smith HJ, Langenberg A, Lekeux P. Measurement of respiratory impedance in healthy calves using the impulse oscillation technique - physiological and methodological aspects. Vet J 1998; 155: 2738. Muller E. Nichtinvasive Differenzierung der mechanischen eigenschaften des respiratorischen systems - strukturermittlung und parameterbestimmung mit hilfe systemtheoretischer verfahren [Noninvasive differentiation of mechanical properties of the respiratory system determination of structures and parameters using methods of system theory]. Dissertation B, Ilmenau, Technische Hochschule Ilmenau, 1989. Oostveen E, Zwart A. Reliability of the coherence function for rejecting respiratory impedance data. Eur Respir Rev 1991; 1: 218221. Miller T, Pimmel R. Standard errors on respiratory mechanical parameters obtained by forced random excitation. IEEE Trans Biomed Eng 1983; 30: 826832. Van Noord JA, Van de Woestijne KP, Demedts M. Clinical applications and modelling of forced oscillation mechanics of the respiratory system. Eur Respir J 1991; 4: 247248. Peslin R, Fredberg J. Oscillation mechanics of the respiratory system. In: Macklem P, Mead J, eds. Handbook of Physiology, Section 3: The respiratory system. Vol III, Mechanics of breathing. Bethesda, American Physiological Society, 1986; pp. 145166. Cauberghs M, Van de Woestijne KP. Changes of respiratory input impedance during breathing in humans. J Appl Physiol 1992; 73: 23552362. Reinhold P, Smith HJ, Close R, Genicot B, Lekeux P. Validation of impulse oscillometry in Friesian and Blue Belgian calves with respect to changes in extrathoracic upper airway resistance. Res Vet Sci 1998; 65: 93101. Ritz T, Dahme B, Dubois AB, et al. Guidelines for mechanical lung function measurements in psychophysiology. Psychophysiology 2002; 39: 546567. Van Noord J, Clement J, Van de Woestijne K, Demedts M. Total respiratory resistance and reactance as a measurement of response to bronchial challenge with histamine. Am Rev Respir Dis 1989; 139: 921926. Bouaziz N, Beyaert C, Gauthier R, Monin P, Peslin R, Marchal F. Respiratory system reactance as an indicator of the intrathoracic airway response to methacholine in children. Pediatr Pulmonol 1996; 22: 713. Duiverman EJ, Neijens HJ, Van der Snee-van Smaalen, Kerrebijn KF. Comparison of forced oscillometry and forced expirations for measuring does-related responses to inhaled methacholine in asthmatic children. Bull Eur Physiopathol Respir 1986; 22: 433436. Landser F, Clement J, Van De Woestijne K. Normal values of total respiratory resis` tance and reactance determined by forced oscillations. Inuence of smoking. Chest 1982; 81: 586591. Wouters EFM, Polko AH, Schouten HJA, Visser BF. Contribution of impedance measurement of the respiratory system to bronchial challenge tests. J Asthma 1988; 25: 259267. Lorino AM, Zerah F, Mariette C, Harf A, Lorino H. Respiratory resistive impedance in obstructive patients: linear regression analysis vs viscoelastic modelling. Eur Respir J 1997; 10: 150155. Daroczy B, Hantos Z. Generation of optimum pseudorandom signals for respiratory impedance measurements. Int J Biomed Comput 1990; 25: 2131. Clement J, Landser F, Van de Woestijne K. Total resistance and reactance in patients with ` respiratory complaints with and without airways obstruction. Chest 1983; 2: 215220. Pasker HG, Schepers R, Clement J, Van de Woestijne KP. Total respiratory impedance measured by means of the forced oscillation technique in subjects with and without respiratory complaints. Eur Respir J 1996; 9: 131139. Cuijpers CEJ, Wesseling G, Swaen GMH, Wouters EFM. Frequency dependence of oscillatory resistance in healthy primary school children. Respiration 1993; 60: 149154.
101
H.J. SMITH ET AL.
39.
40. 41. 42.
43. 44. 45. 46. 47. 48. 49. 50. 51. 52.
53. 54. 55. 56.
57.
58.
59. 60. 61.
Clement J, Dumoulin B, Gubbelmans R, Hendriks S, Van de Woestijne K. Reference values of total respiratory resistance and reactance between 4 and 26 Hz in children and adolescents aged 4 20 years. Bull Eur Physiolpathol Respir 1987; 23: 441448. Solymar L, Aronsson PH, Bake B, Bjure J. Respiratory resistance and impedance magnitude in healthy children aged 2-18 years. Pediatr Pulmonol 1985; 1: 134140. Landser FJ, Nagels J, Clement J, Van de Woestijne KP. Errors in the measurement of total ` respiratory resistance and reactance by forced oscillations. Respir Physiol 1976; 28: 289301. Marotta A, Klinnert M, Price M, Larsen G, Liu A. Impulse oscillometry provides an effective measure of lung dysfunction in 4-year-old children at risk for persistent asthma. J Allergy Clin Immunol 2003; 112: 317322. Mead J. Contribution of compliance of airways to frequency-dependent behaviour of lungs. J Appl Physiol 1969; 26: 670673. Govaerts E, Demedts M, Van de Woestijne KP. Total respiratory impedance and early emphysema. Eur Respir J 1993; 6: 11811185. Goldman M. Clinical application of forced oscillation. Pulm Pharm & Therap 2001; 14: 341350. Imberger H. Elementary complex number analysis of lung models. Respiration 1985; 47: 5769. Mobeireek A, Alhamad A, Al-Subaei A, Alzeer A. Psychogenic vocal cord dysfunction simulating bronchial asthma. Eur Respir J 1995; 8: 19781981. Bucca C, Rolla G, Brussino L, De Rose V, Bugiani M. Are asthma-like symptoms due to bronchial or extrathoracic airway dysfunction? Lancet 1995; 346: 791795. Beraldo PSS, Mateus SRM, Araujo LM, Horan TA. Forced oscillation technique to detect and monitor tracheal stenosis in a tetraplegic patient. Spinal Cord 2000; 38: 445447. Horan T, Mateus S, Bernaldo P, et al. Forced oscillation technique to evaluate tracheostenosis in patients with neurologic injury. Chest 2001; 120: 6973. Uystepruyst C, Reinhold P, Coghe J, Bureau F, Lekeux P. Mechanics of the respiratory system in healthy newborn calves using impulse oscillometry. Res Vet Science 2000; 68: 4755. Goldman M, Carter R, Klein R, Fritz G, Carter B, Pachucki P. Within- and betweenday variability of respiratory impedance using impulse oscillometry in adolescent asthmatics. Ped Pulmonol 2002; 34: 312319. Barnes PJ. Circadian variation in airway function. Am J Med 1985; 79: Suppl. 6a, 59. Bonnet R, Jorres R, Heitman U, Magnussen H. Circadian rhythm in airway responsiveness and airway tone in patients with mild asthma. J Applied Physiol 1991; 71: 15981605. Reinhold P, Uystepruyst C. The intra-subject variability of respiratory impedance - a new parameter of clinical relevance. Eur Respir J 2000; 16: Suppl. 31, 134s. Smolensky M, Barnes P, Reinberg A, McGovern J. Chronobiology and asthma. I. Day-night differences in bronchial patency and dyspnea and circadian rhythm dependencies. J Asthma 1986; 23: 321343. Barnikol W. Perspektiven einer innovativen funktionsdiagnostik des bronchialsystems [Prospectives of innovative function diagnostics of the bronchial system]. Munchen Deisenhofen, Dustri-Verlag, Dr. Karl Feistle, 1997; pp. 2240. Barnikol W, Diether K. Zwei neue kenngroen zur charakterisierung des bronchialen status in vivo und bronchial wirksamer substanzen. Das diskordante verhalten des bronchialsystems [Two new parameters to characterise the bronchial status in vivo and after administration of bronchial active substances. The discordant behaviour of the bronchial system]. Respiration 1980; 39: 97 104. Hafer C, Strohl K, Fouke J. Phasic changes in upper airway impedance. Respir Physiol 1987; 70: 1324. Peslin R, Ying Y, Gallina C, Duvivier C. Within-breath variations of forced oscillation resistance in healthy subjects. Eur Respir J 1992; 5: 8692. Reinhold P, Barnikol W. Breath-by-breath evaluation of the oscillatory resistance as assessed with monofrequency oscillation method in calves - a non-invasive procedure to characterize the motility of the bronchial system. Eur Respir Rev 1994; 4: 178181.
102
62.
63. 64.
65. 66. 67.
68. 69.
70. 71. 72.
73.
74.
75. 76.
77. 78.
79. 80.
81. 82.
Reinhold P, MacLeod D, Lekeux P. Comparative evaluation of impulse oscillometry and a monofrequency forced oscillation technique in clinically healthy calves undergoing bronchochallenges. Res Vet Sci 1996; 61: 206213. Klein C, Reinhold P. Analysis of respiratory mechanics by impulse oscillometry in non-sedated and diazepam-sedated swine. Res Vet Sci 2001; 70: 181189. Klein C, Smith HJ, Reinhold P. Respiratory mechanics in conscious swine: effects of face mask, head position and bronchoconstriction evaluated by impulse oscillometry. Res Vet Sci 2003; 75: 7181. Klein C, Smith H-J. Analysis of respiratory mechanics in sedated and non-sedated horses. Eur Respir J 2003; 22: Suppl. 45, 689. Van Erck E, Votion D, Art T, Lekeux P. Measurement of respiratory function by impulse oscillometry in horses. Equine Vet J 2004; 36: 2128. Van Erck E, Votion D, Kirschvink N, Art T, Lekeux P. Use of the impulse oscillometry system for testing pulmonary function during methacholine bronchoprovocation in horses. Am J Vet Res 2003; 64: 14141420. Marchal F, Loos N. Respiratory oscillation mechanics in infants and preschool children. Eur Respir Mon 1997; 5: 5887. Badia J, Farre R, John Kimoff R, et al. Clinical application of the forced oscillation technique for CPAP titration in the sleep apnea/hypopnea syndrome. Am J Respir Crit Care Med 1999; 160: 15501554. Badia J, Farre R, Montserrat J, et al. Forced oscillation technique for the evaluation of severe sleep apnoea/hypopnoea syndrome: a pilot study. Eur Respir J 1998; 11: 11281134. Farre R, Ferrer M, Rotger M, Torres A, Navajas D. Respiratory mechanics in ventilated COPD patients: forced oscillation versus occlusion techniques. Eur Respir J 1998; 12: 170176. Farre R, Gavela E, Rotger M, Ferrer M, Roca J, Navajas D. Noninvasive assessment of respiratory resistance in severe chronic respiratory patients with nasal CPAP. Eur Respir J 2000; 15: 314319. Farre R, Peslin R, Rotger M, Barbera J, Navajas D. Forced oscillation total respiratory resistance and spontaneous breathing lung resistance in COPD patients. Eur Respir J 1999; 14: 172178. Decramer M, Demedts M, van de Woestijne KP. Isocapnic hyperventilation with cold air in healthy non-smokers, smokers and asthmatic subjects. Bull Eur Physiopathol Respir 1984; 20: 237 243. Nagels J, Landser F, van der Linden L, Clement J, van de Woestijne K. Mechanical properties of ` lungs and chest wall during spontaneous breathing. J Appl Physiol 1980; 49: 408416. Van Noord J, Clement J, Van de Woestijne K, Demedts M. Total respiratory resistance and reactance in patients with asthma, chronic bronchitis and emphysema. Am Rev Respir Dis 1991; 143: 922927. Van Noord J, Smeets J, Clement J, Van de Woestijne K, Demedts M. Assessment of reversibility of airow obstruction. Am J Respir Crit Care Med 1994; 150: 551554. Van Noord JA, Clement J, Cauberghs M, Mertens I, Van de Woestijne KP, Demedts M. Total respiratory resistance and reactance in patients with diffuse interstitial lung disease. Eur Respir J 1989; 2: 846852. Coe C, Watson A, Joyce H, Pride N. Effects of smoking on changes in respiratory resistance with increasing age. Clin Sci 1989; 76: 487494. Phagoo S, Watson R, Silverman M, Pride N. Comparison of four methods of assessing airow resistance before and after induced airway narrowing in normal subjects. J Appl Physiol 1995; 79: 518525. Yap J, Watson R, Gilbey S, Pride N. Effects of posture on respiratory mechanics in obesity. J Appl Physiol 1995; 79: 11991205. Brochard L, Pelle G, de Palmas J, et al. Density and frequency-dependence of resistance in early airway obstruction. Am Rev Resp Dis 1987; 135: 579584.
103
H.J. SMITH ET AL.
83. 84. 85. 86. 87. 88. 89. 90. 91. 92.
93. 94.
95. 96. 97.
98. 99.
100. 101.
102. 103.
104.
Beydon L, Malassine P, Lorino A, et al. Respiratory resistance by end-inspiratory occlusion and forced oscillations in intubated patients. J Appl Phyiol 1996; 80: 11051111. Lorino AM, Hamoudi K, Lofaso F, et al. Effects of continuous negative airway pressure on lung volume and respiratory resistance. J Appl Physiol 1999; 87: 605610. Zerah-Lancner F, Lofaso F, Coste A, Ricol F, Goldenberg F, Harf A. Pulmonary function in obese snorers with or without sleep apnea. Am J Respir Crit Care Med 1997; 156: 522527. Zerah F, Lorino A, Lorino H, Harf A, Macquin-Mavier I. Forced oscillation technique vs. spirometry to assess bronchodilatation in patients with asthma and COPD. Chest 1995; 108: 4147. Peslin R. Methods for measuring total respiratory impedance by forced oscillations. Bull Eur Physiopath Resp 1986; 22: 621631. Peslin R, Duvivier C. Partitioning of airway and respiratory tissue mechanical impedances by body plethysmography. J Appl Physiol 1998; 84: 553561. Peslin R, Duvivier C. Removal of thermal artifact in alveolar pressure measurement during forced oscillation. Respir Physiol 1999; 117: 141150. Peslin R, Duvivier C, Didelon J, Gallina C. Respiratory impedance measured with head generator to minimize upper airway shunt. J Appl Physiol 1985; 59: 17901795. Peslin R, Duvuvier C, Gallina C, Cervantes P. Upper airway artifact in respiratory impedance measurements. Am Rev Respir Dis 1985; 132: 712714. Klug B, Bisgaard H. Measurement of lung function in awake 24-year-old asthmatic children during methacholine challenge and acute asthma: a comparison of the impulse oscillation technique, the interrupter technique, and transcutaneous measurement of oxygen versus wholebody plethysmography. Pediatr Pulmonol 1996; 21: 290300. Hellinckx J, De Boeck K, Bande-Knops J, van der Poel M, Demedts M. Bronchodilator response in 3-6.5 years old healthy and stable asthmatic children. Eur Respir J 1998; 12: 438443. Buhr W, Jorres R, Berdel D, Landser F. Correspondence between forced oscillation and body ` plethysmography during bronchoprovocation with carbachol in children. Pediatr Pulmonol 1990; 8: 280288. Carvalhaes-Neto N, Lorino H, Gallinari C, et al. Cognitive function and assessment of lung function in the elderly. Am J Respir Crit Care Med 1995; 152: 16111615. Wesseling GJ. Respiratory impedance measurements in clinical lung function testing. Proefschrift, Maastricht, April 1993. Sharma G, Arora H, Lester L, Goldman M. Forced oscillatory (IOS) assessment of pulmonary function and bronchodilator response in children with neuromuscular disorders. Am J Respir Crit Care Med 2003; 167: A512. Radulovic M, Spungen A, Wecht J, et al. Effects of neostigmine and glycopyrrolate on pulmonary resistance in spinal cord injury. J Rehabil Res Dev 2004; 41: 5358. Niven A, Backenson T, Goldman M, Weisman I. Resistance values measured by forced oscillation (IOS) are race dependent and increased at low frequencies in non-caucasian subjects. Am J Respir Crit Care Med 2004; 169: A103. Spahn J, Cherniack R, Paull K, Gelfand E. Is forced expiratory volume in one second the best measure of severity in childhood asthma? Am J Respir Crit Care Med 2004; 169: 784786. Saadeh C, Goldman M, Gaylor P, et al. Forced oscillation using impulse oscillometry (IOS) detects false negative spirometry in symptomatic patients with reactive airways. J Allergy and Clin Immun 2003; 111: S136. Hachem R, Chakinala M, Yusen R, et al. The predictive value of bronchiolotis obliterans syndrome stage 0-p. Am J Respir Crit Care Med 2004; 169: 468472. Ross D, Goldman M, Strieter R, Belperio J, Ardehali A. "Multi-frequency forced oscillation technique (FOT) for assessment of lung allograft function: a pilot study". J Heart Lung Transplant 2004; 23: S131. Estenne M, Van Muylem A, Knoop C, Antoine M. Detection of obliterative bronchiolitis after lung transplantation by indexes of ventilation distribution. Am J Respir Crit Care Med 2000; 162: 10471051.
104
105. Quanjer PhH, Tammeling GJ, Cotes JE, Pedersen OF, Peslin R, Yernault JC. Lung volume and forced ventilatory ows. Report Working Party, Standardization of lung function tests. European Coal and Steel Community. Ofcial statement of the European Respiratory Society. Eur Respir J 1993; 6: Suppl. 16, 540. 106. Orehek J, Nicoli MM, Delpierre NS, Beaupre A. Inuence of the previous deep inspiration in the spirometric measurements of provoked bronchoconstriction in asthma. Am Rev Respir Dis 1981; 123: 569572. 107. Salome CM, Thorpe CW, Diba C, Brown NJ, Berend N, King GG. Airway re-narrowing following deep inspiration in asthmatic and nonasthmatic subjects. Eur Respir J 2003; 22: 6268. 108. Schmekel B, Smith HJ. The diagnostic capacity of forced oscillation and forced expiration techniques in identifying asthma by isocapnic hyperpnoea of cold air. Eur Respir J 1997; 10: 2243 2249. 109. Jackson AC, Gulesian PJ Jr, Mead J. Glottal aperture during panting with voluntary limitation of tidal volume. J Appl Physiol 1975; 39: 834836. 110. Smith T, DuBois A. The effects of scopolamine on the airways of man. Anesthesiology 1969; 30: 1218. 111. Hensley MJ, OCain CF, McFadden ER Jr, Ingram RH Jr. Distribution of bronchodilatation in normal subjects: beta agonist versus atropine. J Appl Physiol 1978; 45: 778782. 112. Bankier AA, Van Muylem A, Knoop C, Estenne M, Gevenois PA. Bronchiolitis obliterans syndrome in heart-lung transplant recipients: diagnosis with expiratory CT. Radiology 2001; 218: 533539. 113. Nielsen K, Bisgaard H. The effect of inhaled budesonide on symptoms, lung function, and cold air and methacholine responsiveness in 2- to 5-year-old asthmatic children. Am J Respir Crit Care Med 2000; 162: 15001506. 114. Nielsen K, Bisgaard H. Lung function response to cold air challenge in asthmatic and healthy children of 2-5 years of age. Am J Respir Crit Care Med 2000; 161: 18051809. 115. Nielsen K, Bisgaard H. Discriminative capacity of bronchodilator response measured with three different lung function techniques in asthmatic and healthy children aged 2 to 5 years. Am J Respir Crit Care Med 2001; 164: 554559. 116. Grimby G, Takishima T, Graham W, Macklem P, Mead J. Frequency dependence of ow resistance in patients with obstructive lung disease. J Clin Invest 1968; 47: 14551465. 117. Kjeldgaard J, Hyde R, Speers D, Reichert W. Frequency dependence of total respiratory resistance in early airway disease. Am Rev Respir Dis 1976; 144: 501508. 118. Goldman M, Knudson R, Mead J, Peterson N, Schwaber J, Wohl M. A simplied measurement of respiratory resistance by forced oscillations. J Appl Physiol 1970; 28: 113116. 119. Bhansali P, Irvin C, Dempsey J, Bush R, Webster J. Human pulmonary resistance: effect of frequency and gas physical properties. J Appl Physiol 1979; 47: 161168. 120. Niewoehner DE, Kleinerman J, Rice DB. Pathologic changes in the peripheral airways of young cigarette smokers. N Engl J Med 1974; 291: 755758. 121. Frantz I, Close R. Alveolar pressure swings during high frequency ventilation in rabbits. Pediatr Res 1985; 19: 162166. 122. Malmberg LP, Pelkonen A, Poussa T, Pohianpalo A, Haahtela T, Turpeinen M. Determinants of respiratory system input impedance and bronchodilator response in healthy Finnish preschool children. Clin Physiol Funct Imaging 2002; 22: 6471. 123. Ortiz G, Menendez R. The effects of inhaled albuterol and salmeterol in 2- to 5-year-old asthmatic children as measured by impulse oscillometry. J Asthma 2002; 39: 531536. 124. Nielsen K, Bisgaard H. Lung function response to cold air challenge in asthmatic and healthy children of 2-5 years of age. Am J Respir Crit Care Med 2000; 161: 18051809. 125. Ward P, Hunninghake G. Lung inammation and brosis. Am J Respir Crit Care Med 1998; 157: S123S129. 126. Weibel ER. Morphometry of the Human Lung. New York, Academic Press 1963.
105

Monograph Forced Oscillation Technique

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Monograph Forced Oscillation Technique

Uploaded by

Copyright:

Available Formats

CHAPTER 5

Forced oscillation technique and impulse oscillometry

Methodology of impulse oscillation technique

OSCILLOMETRY: FOT AND IOS

Peculiarities of aperiodic waveforms

H.J. SMITH ET AL.

Rrs 0.5 0.4 kPasL-1 0.3 0.2

Xrs 0.3 0.2 0.1 0

Coherence g2 () Regional nonhomogeneity Resistance Rrs ()

g2 1.0 0.9 0.8 0.7 0.6 0.5

res Reactance Xrs() Reactance area (AX)

0.1 -0.1 0 -0.2

5 0 Attenuation dB -5 -10 -15 -20 -25

OSCILLOMETRY: FOT AND IOS

Impulse oscillometry system

Impulse generator Y-adapter

Flow transducer Mouthpiece Pressure transducer

H.J. SMITH ET AL.

OSCILLOMETRY: FOT AND IOS

a) 1.00 Amplitude of primary flow Ls-1 0.75 0.50 0.25 0.00

Approximated baseline Pulserate 5s-1 200 ms

-0.25 -0.50 0.0 0.1 0.2

b) Corrected impulse flow Ls-1

0.3 0.4 Recording time s

Resulting impulse pressure

Corrected impulse pressure Pa

H.J. SMITH ET AL.

Interpretation of oscillation mechanics

OSCILLOMETRY: FOT AND IOS

H.J. SMITH ET AL.

Capacitive spectral range of reactance. Thoraco-pulmonary elasticity is commonly

OSCILLOMETRY: FOT AND IOS

Capacitance versus dynamic pulmonary compliance. Dynamic pulmonary compliance

H.J. SMITH ET AL.

Inertive spectral range of reactance. During resting breathing at normal respiratory

Variability of oscillometric parameters

Methodological versus biological variability. Low methodological variability in

Special application of impulse oscillometry system to animals

OSCILLOMETRY: FOT AND IOS

Relevance of oscillometry in clinical practice

H.J. SMITH ET AL.

OSCILLOMETRY: FOT AND IOS

H.J. SMITH ET AL.

3 Volume L Zrs5 kPasL-1 2 1 0 -1 2.0 1.6 1.2 0.8 0.0

0.4 0 10 20 30 40 Recording time s 50 60 70

OSCILLOMETRY: FOT AND IOS

Oscillometry in relationship to other diagnostic pulmonary functional tests

Oscillometry as a clinical monitor of response to treatment

H.J. SMITH ET AL.

Oscillometry in the clinical pulmonary laboratory

Obstructive lung disease

OSCILLOMETRY: FOT AND IOS

Resistance Rrs kPasL-1

H.J. SMITH ET AL.

1.0 0.8 0.6 0.4 0.2 0.0

Resistance Rrs kPasL-1

Central airway obstruction. Because of the frequency-dependent distribution of

OSCILLOMETRY: FOT AND IOS

Nonresistive components of forced oscillation technique. A second discriminant

Resistance Rrs kPasL-1

H.J. SMITH ET AL.

OSCILLOMETRY: FOT AND IOS