Which position is most favorable position for performing the assessment of breath sounds?

Abstract

Auscultation is defined as “the act of listening for sounds within the body, chiefly for ascertaining the condition of the lungs, heart, pleura, abdomen, and other organs.”1(p139) Through auscultation, physical therapists look for signs of excessive secretions (presence of added sounds) or obtain evidence of satisfactory lung inflation (satisfactory “air entry”). In clinical circumstances, lung sound intensity is often related to lung volume, and an increase in lung sound intensity on auscultation is considered indicative of lung expansion.

In a side-lying position, the weight of the abdominal contents pushes the dome of the lower diaphragm farther into the thorax than does the dome of the upper diaphragm.2(p102) Despite a diminished lung volume, the lower diaphragm contracts more effectively than the upper diaphragm during inspiration and, therefore, ventilation distributes preferentially to the dependent lung.2(p141) There is evidence that breath sound intensity correlates with pulmonary ventilation.3 Based on this evidence, lung sound intensity should be higher in the dependent lung than in the nondependent lung. There are relatively few reports, however, examining the effect of positioning on lung sound intensity.

We investigated the intensity and spectral characteristics of lung sounds recorded electronically in the sitting and side-lying positions in young adults without pulmonary dysfunction. We hypothesized that there would be no differences in the data recorded in corresponding regions between (1) the left and right lungs in the sitting position, (2) the dependent and nondependent lungs in the side-lying position (in side-lying positions, the upper hemithorax is “nondependent,” and the side in contact with the bed is “dependent”), (3) the sitting position and the dependent position, or (4) the sitting position and the nondependent position.

Method

Subjects

Physical therapist students were invited (by an open letter displayed on the students' bulletin board) to participate in a session in which lung sounds would be recorded electronically. At the session, before examination, students were asked whether they were free from any known cardiorespiratory disease. Each subject was then asked to perform spirometry testing, using a Microspiro HI-298 spirometer.* Subjects with any one of the following lung function variables that were less than 85% of the predicted value were excluded from the study: forced vital capacity, forced expiratory volume in 1 second, and peak expiratory flow. Thirteen students responded to the invitation letter. Two respondents (one who had asthma and one who had a cold with audible crackles) were excluded from the study. Data from the remaining 11 subjects (5 male, 6 female) were used in the study. The subjects had a mean age of 20.4 years (SD=2.7, range=19–26), a mean height of 166.3 cm (SD=6.5, range=154–175), and a mean weight of 57.5 kg (SD=10, range=45–75).

Instrumentation

Three identical microphones (Realistic Electret 33-1052†) were attached to 3 stethoscope chest pieces (Littmann Classic II‡) via 5-cm-long rubber tubing. Signals from the microphone-stethoscope chest pieces were samples at a frequency of 5 kHz and amplified by 3 separate amplifiers (Bruel & Kjaer type 2609§). Output from the amplifier was converted to a digital signal by aPCMCIA A/D PC card (DT7102 PC Card-EZ‖) via a multichannel connector panel (DT784‖). The PC card was installed in a notebook computer (Compaq Presario 1082 with Intel Pentium 166-MHz CPU, 48MB RAM#) (Fig. 1). Signals from a pneumotachograph** were directed simultaneously to an oscilloscope (Philips PM3350A††) and to one of the channels of the connector panel via a pressure transducer (Bio Precision PT-010‡‡). The pneumotachograph was calibrated before each sound recording session with a 2-L syringe, such that a flow rate of 1 L/s was equivalent to 0.1 V.).

Figure 1

Diagrammatic presentation of the equipment setup for the study.

Equipment Calibration

All microphones were calibrated using a sound level (Bruel & Kjaer type 4231§) and had a sensitivity of 3 mV/Pa. The performance of each microphone-stethoscope chest piece was compared with that of a reference standard microphone (Bruel & Kjaer type4166§), based on US Bureau of National Standards data. The calibration showed a coherence function value between 0.9 and 1 (frequency range=31.5–500 Hz) and a negligible, in terms of this study, performance variation of 1% to 3%.

Experimental Procedure

The 3 chest-piece diaphragms were attached to the left thigh (background sound) and the left and right posterior chest walls with three 4- × 10-cm Hypafix adhesive dressing retention sheets,§§ reinforced by three 1.5- × 10-cm strips of Micropore surgical tape‡ (Fig. 2). The posterior chest wall attachment was at the eighth intercostal space in the midscapular line.

Figure 2

Stethoscope chest piece attachment.

Each subject was seated in front of an oscilloscope with his or her forearms supported and both hands holding the pneumotachograph (Fig. 3). A noseclip was fitted to ensure mouth breathing during sound recording. The oscilloscope displayed signals on-line from the pneumotachograph to provide visual feedback of the pattern of each breath to the subject. The subject was asked to practice 6-second breathing cycles with a peak inspiratory flow rate of approximately 1.5 to 2 L/s (a flow rate also used by other researchers4). A standardized flow rate has been used as lung sound spectra vary at different airflows.4 Our experience shows that tidal breathing in young subjects falls within this range. Two horizontal grid lines (±0.2 V) were highlighted on the oscilloscope screen. The subject was then given the following instruction: “Take a breath in and out through your mouth, using 3 seconds for breathing in and 3 seconds for breathing out, trying to keep your breathing curve close to, but within, the 2 identified lines on this screen.”

Figure 3

Subject during lung sound recording in the sitting position.

Lung sounds for 5 breaths were recorded with the HPVEE for Windows software program.‖‖ This procedure was repeated with the subject in the right side-lying position and then in the left side-lying position. In the side-lying positions, the subject lay on a plinth with a pillow under the head and with a pillow between the knees, which were slightly flexed. The angle of hip and knee flexion was not standardized, but the subjects were reminded that they should be comfortable and able to relax in the position. The elbows were flexed, with the hands holding the pneumotachograph as in the sitting position. The oscilloscope was positioned so that the subject could view the screen.

To minimize data variability as a consequence of the recording apparatus, particularly in an attempt to limit any difference in data recorded being due to a difference in performance of the 2 chest pieces, the left and right chest pieces were interchanged after recordings from 5 subjects were obtained.

Data Processing

The raw data were visually examined before processing. Data were discarded if the peak inspiratory flow signal was less than 0.1 V or greater than 0.2 V or the data appeared clipped (ie, a recorded sound intensity beyond the processing capabilities of the hardware). The inspiratory and expiratory phases of each breath were identified from the highest and lowest points in the flow signal curve using the positive and negative flow signals from the pneumotachograph. A signal window that included 512 data points to the left and right of the highest data point was subjected to fast Fourier transform analysis (a numerical algorithm for fast computation of discrete signals from time domain to frequency domain). The inspiratory frequency spectrum obtained from the “background” chest piece was subtracted from the inspiratory spectrum recorded from the left or right chest piece. The spectra from 5 breaths were averaged to provide a resultant inspiratory or expiratory spectrum. Positional data in the same subject were analyzed using paired t tests with the SPSS for Windows, version 7.5, statistical package.## The P <.05 level was considered a statistically significant difference between the left and right lung data and the sitting and side-lying positions over the same hemithorax.

Variables recorded for analysis were peak intensity (the maximum power measured from processed spectral sound data), frequency at maximum power (the frequency at which sound amplitude is highest), and median frequency (the frequency at which the spectrum energy is halved). These are common variables used to characterize signals as a function of frequency. Peak intensity corresponds to the sound intensity heard through the stethoscope. Frequency at maximum power provides information regarding the frequency range of the sound activity. This information is more useful when added sounds, such as crackles and wheezes, are investigated. Median frequency is useful for describing a signal pattern, but it is not directly related to any specific clinical observation.

Results

Figure 4 shows a typical raw data file recorded from one subject in the right side-lying position. Figure 5 illustrates the averaged and smoothed spectral curves of the same subject recorded during inspiration and expiration. Inspiratory lung sound intensity was louder than expiratory sounds in both the right and left lungs (P <.02) (Tab. 1).

Table 1

Comparison of Sound Data Recorded From the Left and Right Chest Wall at Different Positionsa

a

Data are mean (SD). Asterisk (*) denotes P<.05 (paired t test, df=10). Box denotes data recorded with the lung in the dependent position.

Table 1

Comparison of Sound Data Recorded From the Left and Right Chest Wall at Different Positionsa

a

Data are mean (SD). Asterisk (*) denotes P<.05 (paired t test, df=10). Box denotes data recorded with the lung in the dependent position.

Figure 4

Raw data file recorded from one of the subjects in the right side-lying position.

Figure 5

Averaged and smoothed spectral curves of the same subject during inspiration and expiration.

Left and Right Lungs in the Sitting Position

In the sitting position, the peak intensity of the inspiratory sound recorded over the left chest wall was found to be higher than that recorded over the right chest wall (25.6 and 20.7 dB, respectively) (P<.05) (Tab. 1). Variables recorded during expiration over the right and left sides in the sitting position were not different (Tab. 1).

Dependent and Nondependent Positions

In the left side-lying position, the peak intensity recorded during inspiration from the dependent (left) side was higher than that recorded from the nondependent (right) side (23.5 and 15.7 dB, respectively) (P <.0005). In the right side-lying position, a similar result was obtained only during expiration (expiratory sound intensity was 11.2 dB for the right lung and 6.8 dB for the left lung) (P <.01).

Sitting and Nondependent Positions

The peak sound intensity was greater in the sitting position than in the nondependent side-lying position in the left and right lungs during inspiration (P <.05) (Tab. 2). For the left lung, the sound intensity recorded in the sitting position was 25.6 dB, and the sound intensity recorded in the nondependent side-lying position was 19.7 dB. For the right lung, the sound intensity recorded in the sitting position was 20.7 dB, and the sound intensity recorded in the nondependent side-lying position was 15.7 dB. The frequency at maximum power recorded over the right lung was also higher in the sitting position (244.5 Hz) than in the nondependent side-lying position (201.6 Hz) (P <.005). There was no difference in variables between the sitting position and the nondependent side-lying position during expiration.

Table 2

Comparison of Sound Data Recorded Over the Same Lung in the Sitting and Side-lying Positionsa

a

Data are mean (SD). Asterisk (*) denotes P<.05 (paired t test, df=10).

Table 2

Comparison of Sound Data Recorded Over the Same Lung in the Sitting and Side-lying Positionsa

a

Data are mean (SD). Asterisk (*) denotes P<.05 (paired t test, df=10).

Sitting and Dependent Positions

Over the left posterior chest wall, there were no differences in any variables between the sitting and left side-lying positions during either inspiration or expiration (Tab. 2). On the right side during inspiration, the frequency at maximum power was higher in the side-lying position (278.4 Hz) than in the sitting position (244.5 Hz) (P <.05).

Discussion

In clinical practice, it is customary to compare lung sound intensity in one area over the chest wall with that of the corresponding area on the opposite side.5 Our study demonstrated that in the sitting position, the inspiratory lung sound recorded over the left posterior chest wall was louder than that recorded over the corresponding position on the right. Lung sound analysis has shown that not only was there marked intersubject and intrasubject variation but bilateral symmetry in lung sound amplitude was also absent.6 Kraman7 demonstrated that sound intensity at one base (left or right) was greater than at the opposite base in 7 of 9 subjects and that sound intensity at the left apex was always louder than, or equal to, that at the right apex. A recent study by Pasterkamp and coworkers8 also demonstrated that breath sounds were louder on the left than on the right at the posterior lung base. As pointed out by Pasterkamp and colleagues, the observed differences in sound intensity at posterior bases are very likely due to the asymmetry of the geometry of airways in both lungs. Because the left bronchus is smaller but more horizontal than the right bronchus and because of the position of the heart, many of the major left segmental bronchi are directed more posteriorly compared with the right bronchi. The findings of our study suggested that clinicians who assess lung sounds should be aware of the differences in lung sound intensity that routinely occur between the left and right bases.

In the side-lying positions, the sound intensity recorded over the dependent lung was louder than that recorded over the nondependent lung. The differences measured in the right side-lying position were smaller than those measured in the left side-lying position and probably require a larger sample size for statistical evidence to be supported. Only one other study9 has analyzed breath sounds in the side-lying position. In that study, Leblanc and colleagues demonstrated that at a lung volume higher than residual volume, maximal breath sound intensity took place in the inferior part of the lung in the lateral position. They concluded that breath sound intensity was a good sign of regional pulmonary ventilation. We standardized each subject's inspiratory flow to about 2 L/s so that our data would be comparable to that of previously published work. The inspiratory flow for normal tidal breathing in young individuals without pulmonary dysfunction falls within this range. This inspiratory flow should result in a lung volume of at least 30% of vital capacity, which, in young subjects, is the lung volume at which all airways are open and when breath sound intensity corresponds to pulmonary ventilation.9 If the inspired volume is greater than 30% of vital capacity, then air entry will be greater in the dependent lung, but if lung volume is less than 30% of vital capacity, then air will be distributed to the nondependent lung.9

We believe that the relationship between vital capacity and air distribution may explain why the lung sound intensity often appears to be lower in the dependent lung than in the nondependent lung during auscultation of a patient in the side-lying position. Auscultation of the lungs in a side-lying position is often performed on patients who are weak and incapable of maintaining a sitting position. The inspiratory effort made by this type of patient is often poor, with a consequently low inspired lung volume. If this lung volume is less than 30% of vital capacity, air will be preferentially distributed to the nondependent lung and result in a lower lung sound intensity in the dependent lung. If a patient with frailty is asked to inspire more fully, lung sound intensity in the dependent lung could, theoretically, improve to the extent that it is higher than that of the nondependent lung. Lung sound intensity was shown to correlate with pulmonary ventilation,3 but inspiratory flow rate and lung volume must also be considered. The same study3 showed that regional breath sounds heard through the stethoscope, without taking lung volume into consideration, correlate poorly with regional ventilation.

Due to the effect of gravity, resting lung volume is higher in the nondependent lung than in the dependent lung. To assist in reexpansion of collapsed or partially collapsed lung segments, the physical therapist may position a patient so that the collapsed lung segment is in a nondependent position. If lung sound intensity is to be used as a variable to indicate lung reexpansion, clinicians should be aware of whether a comparison of this variable is made before and after a technique is applied to the same lung or whether a comparison is made relative to the “nonaffected” but dependent lung.

The sound intensity in both the left and right lungs in the sitting position was greater than over the same lung segments in a nondependent position. There was no difference, however, in lung sound intensity between the sitting position and a dependent side-lying posture when the same lung segment was considered. These findings perhaps can also be explained by better ventilation of the dependent part of the lungs in the sitting and side-lying positions. Clinically, it follows that overall ventilation of both lungs is better in the sitting position than in either side-lying position. The patient has a mechanical advantage in the sitting position, as compared with the side-lying posture. The sitting position results in a deeper breath for the same amount of muscular effort and consequently an increased inspiratory air flow. Thus, inspiratory breath sound intensity is enhanced in the sitting position.

We designed our study to reflect clinical situations. Lung sound analysis allows quantitative comparison of variables such as peak intensity, frequency at maximum power, and median frequency over a long time frame.

These variables are commonly used to describe the characteristic frequency pattern of sound signals. Lung sound measurement is difficult in practice because the equipment setup is complicated, even though advances in computing have greatly simplified data acquisition and analysis. In our study, we used stethoscope chest pieces rather than microphone shrouds (housings) to mimic the clinical setting. We used a sound subtraction technique to remove background noise. We chose to use each subject as his or her own control to limit intersubject variability, and we did not use high-pass filtering because of the spectral overlap between lung and heart sounds.10 Using a technique described by other workers11 to limit the effect of air flow on lung sound spectra, we provided a real-time visual display of peak inspiratory flow rate on an oscilloscope to encourage a repeatable breathing pattern.4,12,13 One limitation of our study relative to application to clinical practice was our use of young subjects without pulmonary dysfunction instead of patients with any impairment of the cardiopulmonary system.

Aging has effects on both anatomical and physiological function of the respiratory system. Tidal breathing in young people without pulmonary dysfunction can reach a lung volume that results in ventilation being preferential to the dependent lung. The effect of aging (or lung pathology) on the lung and chest wall compliance may induce remarkable changes in the distribution of ventilation.14(pp98–101) In pathology where the dependent lung is no longer preferentially better ventilated, or if the lung volume is restricted, the lung sound intensity will decrease accordingly. Therefore, when auscultating elderly patients or patients with chronic respiratory problems, they should be encouraged to inspire using a higher flow or to inspire to a relatively deeper lung volume to permit appropriate assessment of the quality of lung sound.

The shifting of mediastinal structures is minimal in young people without pulmonary dysfunction, but this shifting may have a great effect on airflow and lung volume in patients with cardiopulmonary pathology. In a side-lying position where mediastinal structures are compressed, decreased ventilation will affect the sound intensity accordingly. Research in this area to provide evidence on the effect of compressed mediastinal structures on sound transmission is necessary, as is research on more clinically relevant groups of subjects.

The statement “pulmonary ventilation is preferentially better in the dependent lung” appears in most respiratory physiology books,2(pp142–143),14(p20) but before a physical therapist positions a patient with his or her “affected” lung in the dependent position (be it for maximizing ventilation or arterial saturation), the pathology of the patient, the compliance of the chest wall, and the ultimate goal of the treatment plan should be considered. Similarly, we believe practitioners must not unquestioningly relate lung sound intensity to lung expansion. They should consider the patient's age, pathology, and posture as well as the lung volume and inspiratory flow rate when the lung sounds were auscultated.

Although auscultation is commonly used for the assessment of the respiratory system, the interrater reliability for auscultation of both tape-recorded lung sounds15–17 and patient breath sounds18 was found to be poor. Advances in computer technology have greatly facilitated lung sound analysis, crackles can now be counted, and wheezes with varying frequencies and durations in different disease states can be analyzed quantitatively. Although the stethoscope has changed very little since it was developed more than 160 years ago, this does not mean that electronic auscultation can replace auscultation with a conventional stethoscope. Conventional auscultation has the advantages of being convenient, efficient, and inexpensive. Health care practitioners, however, should be aware of the influence of “artifacts” on auscultatory findings.

Conclusion

For a defined inspiratory flow, in the sitting position, inspiratory lung sound recorded over the left posterior chest wall was louder than the sound recorded over a corresponding area on the right side. In the side-lying positions, the sound intensity was louder over the dependent lung than over the nondependent lung only in the left side-lying position. For both lungs, the sound electronically recorded in the sitting position was higher than that recorded in the same lung segment in a nondependent position, but there was no difference in sound intensity between the sitting position and the dependent side-lying posture.

Concept and research design were provided by A Jones, RD Jones, Kwong, and Burns; writing, by A Jones and RD Jones; data collection, by A Jones; data analysis, by A Jones, RD Jones, and Kwong; project management, A Jones and Burns; fund procurement, by RD Jones and Kwong; subjects, by A Jones; facilities and equipment, by A Jones and Kwong; institutional liaisons, by A Jones; clerical/secretarial support, by RD Jones; and consultation (including review of manuscript prior to submission), by RD Jones, Kwong, and Burns. Kevin Dowsey, Technical Director, Total Turnkey Solutions, provided technical support in the preparation of software for sound data acquisition and analysis. Sik-Cheung Siu, Senior Technician, Department of Rehabilitation Sciences, The Hong Kong Polytechnic University, provided technical assistance during the study.

This study was approved by the Departmental Research Committee, Department of Rehabilitation Sciences, The Hong Kong Polytechnic University.

This project was jointly supported by funding from The Hong Kong Polytechnic University Research Grants and Division of Anaesthesiology and Intensive Care, University of Queensland.

*

Chest Corporation, 3-6-10, Hongo, Bunkyo-ku, Tokyo 113, Japan.

†

Intertan Australia Ltd, 91 Kurrajong Ave, Mt Druitt, New South Wales, 2770 Australia.

‡

3M, Health Care Division, St Paul, MN 55144.

§

Bruel & Kjaer, World Headquarters, DK-2850 Naerum, Denmark.

‖

Data Translation Inc, 100 Locke Dr, Marlboro, MA 01752.

#

Compaq Computer Corp, 20555 FM 149, Houston, TX 77070.

**

Hans-Rudolph Inc, 7200 Wyandotte, Kansas City, MO 64114.

††

Philips Oscilloscope, Fluke Europe BV, PO Box 1186, 502 BD Dindihoven, the Netherlands.

‡‡

Distributed by Hans-Rudolph Inc, 7200 Wyandotte, Kansas City, MO 64114.

§§

Smith & Nephew, Laboratories FISCH, France.

‖‖

Hewlett-Packard Co, 815 SW 14th St, Loveland, CO 80537.

##

SPSS Inc, 444 N Michigan Ave, Chicago, IL 60611.

References

1

Dorland's Illustrated Medical Dictionary

.

26th ed

.

Philadelphia, Pa

:

WB Saunders Co

;

1981

.

2

Nunn

JF

Applied Respiratory Physiology

. 3rd ed.

London, England

:

Butterworth & Co (Publishers) Ltd

;

1987

.

3

Ploy-Song-Sang

Y

Breath sounds and regional ventilation

.

Am Rev Respir Dis

.

1977

;

116

:

187

–

199

.

4

Malmberg

LP

Significant differences in flow standardised breath sound spectra in patients with chronic obstructive pulmonary disease, stable asthma, and healthy lungs

.

Thorax

.

1995

;

50

:

1285

–

1291

.

5

Hinshaw

HC

Diseases of the Chest

.

Philadelphia, Pa

:

WB Saunders Co

;

1980

:

17

–

24

.

6

Dosani

R

Lung sound intensity variability in normal men: a contour phonopneumograhic study

.

Chest

.

1983

;

83

:

628

–

631

.

7

Kraman

S

Lung sounds: relative sites of origin and comparative amplitudes in normal subjects

.

Lung

.

1983

;

161

:

57

–

64

.

8

Pasterkamp

H

Asymmetry of respiratory sounds and thoracic transmission

.

Med Biol Eng Comput

.

1997

;

35

:

103

–

106

.

9

Leblanc

P

Breath sounds and distribution of pulmonary ventilation

.

Am Rev Respir Dis

.

1970

;

102

:

10

–

16

.

10

Ploy-Song-Sang

Y

Characteristics of normal lung sounds after adaptive filtering

.

Am Rev Respir Dis

.

1989

;

139

:

951

–

956

.

11

Gavriely

N

Spectral characteristics of chest wall breath sounds in normal subjects

.

Thorax

.

1995

;

50

:

1292

–

1300

.

12

Pasterkamp

H

Lung sound spectra at standardized air-flow in normal infants, children, and adults

.

Am J Respir Crit Care Med

.

1996

;

154

:

424

–

430

.

13

Schreur

HJW

Lung sound intensity in patients with emphysema and in normal subjects at standardised airflows

.

Thorax

.

1992

;

47

:

674

–

679

.

14

West

JB

Respiratory Physiology: The Essentials

. 5th ed.

Baltimore, Md

:

Williams & Wilkins

;

1995

:

20

,

98

–

101

.

15

Allingame

S

Accuracy and reliability of physiotherapists in the interpretation of tape-recorded lung sounds

.

Australian Journal of Physiotherapy

.

1995

;

41

:

179

–

184

.

16

Aweida

D

Accuracy and reliability of physical therapists in auscultating tape-recorded lung sounds

.

Physiotherapy Canada

.

1990

;

42

:

279

–

282

.

17

Brooks

D

Accuracy and reliability of “specialized” physical therapists in auscultating tape-recorded lung sounds

.

Physiotherapy Canada

.

1993

;

45

:

21

–

24

.

18

Brooks

D

Interrater reliability of auscultation of breath sounds among physical therapists

.

Phys Ther

.

1995

;

75

:

1082

–

1088

.

© 1999 American Physical Therapy Association

© 1999 American Physical Therapy Association

What is respiratory assessment position?

Which of the following is the most comfortable position for a client with a respiratory problem?

What is the best position in assessing the thorax and lungs?