Atelectasis

Full Terms & Conditions of access and use can be found at http://www.tandfonline.com/action/journalInformation?journalCode=iptp20

Physiotherapy Theory and Practice An International Journal of Physical Therapy

ISSN: 0959-3985 (Print) 1532-5040 (Online) Journal homepage: http://www.tandfonline.com/loi/iptp20

Comparison of breathing patterns, pressure, volume, and flow characteristics of three breathing techniques to encourage lung inflation in healthy older people

Chulee Ubolsakka-Jones PhD, PT, Wiraporn Tasangkar MSc, PT & David A Jones PhD

To cite this article: Chulee Ubolsakka-Jones PhD, PT, Wiraporn Tasangkar MSc, PT & David A Jones PhD (2018): Comparison of breathing patterns, pressure, volume, and flow characteristics of three breathing techniques to encourage lung inflation in healthy older people, Physiotherapy Theory and Practice, DOI: 10.1080/09593985.2018.1477890

To link to this article: https://doi.org/10.1080/09593985.2018.1477890

Published online: 25 May 2018.

Submit your article to this journal

Article views: 1

View related articles

View Crossmark data

REPORT

Comparison of breathing patterns, pressure, volume, and flow characteristics of three breathing techniques to encourage lung inflation in healthy older people Chulee Ubolsakka-Jones, PhD, PTa, Wiraporn Tasangkar, MSc, PTb, and David A Jones, PhDc

aSchool of Physical Therapy, Faculty of Associated Medical Sciences, Khon Kaen University, Khon Kaen, Thailand; bPhysical Therapy Department, Bumrungrad International Hospital, Bangkok, Thailand; cSchool of Health Care Sciences, Manchester Metropolitan University, Manchester, United Kingdom

ABSTRACT Background: It is important to encourage lung inflation to prevent postsurgical pulmonary complications and we compared three breathing techniques that place different emphasis on inspiratory flow and breath-holding. Methods: Fourteen healthy older people (69 ± 3.6 yrs) used diaphragmatic breathing (DB), Triflo II (TF), and a water pressure threshold device (BreatheMAX; BM) in a randomized and balanced crossover design. Outcome measures were inspiratory flow and pressure, inspiratory time (Ti), tidal volume (Vt), and breathing frequency. Results: Inspiratory flow with TF was significantly faster than DB and BM (p < 0.001: 0.96 ± 0.1; 0.43 ± 0.20 and 0.28 ± 0.1 L.s−1, respectively) and pressures greater (p < 0.001: −1.3 ± 0.6, −5.5 ± 1.2 and −2.8 ± 3.6 cm H2O). However, Ti was shorter (TF, 1.16 ± 0.21s; DB, 3.31 ± 0.97 s, p < 0.001; BM, 5.53 ± 1.92 s, p < 0.001), resulting in smaller Vt (TF, 1.12 ± 0.29 L; DB, 1.28 ± 0.29L, p = 0.003; BM, 1.37 ± 0.43L, p = 0.016). Breathing frequency was faster with TF compared to DB and BM (p < 0.001). Conclusions: Substantial lung inflation could be achieved with any of the above- mentioned methods, although Vt was smaller with TF and the high inspiratory flow with this method may not inflate the lower lung. The high pressures and rapid breathing with TF could increase the sense of effort. Trials are needed to determine the clinical value of the different breathing exercises.

ARTICLE HISTORY Received 12 March 2017 Revised 18 January 2018 Accepted 4 February 2018

KEYWORDS Incentive spirometry; lung inflation; breathing pattern

Introduction

Postoperative pulmonary complications resulting in atelec- tasis are common causes of morbidity and mortality in more than 75% of patients receiving a neuromuscular blocking agent (Branson, 2013; Miskovic and Lumb, 2017), making a major contribution to the health-care costs (Branson, 2013). The basal lung is the most affected region in 90% of patients with atelectasis following intuba- tion (Poelaert, Szegedi, and Blot, 2013), in the first few days following upper abdominal and chest surgery, with older patients being especially at risk (Johnson et al., 2007).

Incentive spirometry (IS) may be used as a part of a package of techniques to encourage lung expansion (Branson, 2013), which includes the mobilization of secre- tions and early ambulation, together with coughing, oral care, and elevation of the bed head together with adequate analgesia (Cassidy et al., 2013; Kelkar, 2015; Lawrence, Cornell, and Smetana, 2006). However, the value of chest physiotherapy in general, and particularly the role of IS, has been questioned (Do Nascimento Junior et al., 2014;

Freitas, Soares, Cardoso, and Atallah, 2012; Jenkins et al., 1989). Part of the problem may be that the term “incentive spirometer” includes a variety of devices that target differ- ent aspects of breathing, some of which may not encourage the patient to adopt, or continue with, a breathing pattern that helps inflate the lower lung.

The IS devices are mainly classified into those that provide direct feedback of the inspiratory volume (i.e., volume-oriented incentive spirometer [VIS]), which are used to encourage lung inflation and those that provide feedback of flow (i.e., flow-oriented incentive spirometer [FIS]). There have been a number of studies comparing VIS devices, such as Voldyne and Coach, with FIS devices such as Triflo II, Cliniflo and Respirex, the general conclu- sion being that VIS devices lead to greater abdominal expansion (Lunardi et al., 2014; Paisani et al., 2013; Tomich et al., 2007), implying greater diaphragm excursion and expansion of the lower lung. However, despite the well-documented limitations of FIS (particularly Triflo) devices, they are widely used. While almost any device

CONTACT Chulee Ubolsakka-Jones, PhD, PT [email protected] School of Physical Therapy, Faculty of Associated Medical Sciences, Khon Kaen University, Khon Kaen, Thailand. Color versions of one or more figures in the article can be found online at www.tandfonline.com/iptp.

PHYSIOTHERAPY THEORY AND PRACTICE https://doi.org/10.1080/09593985.2018.1477890

© 2018 Taylor & Francis

could be useful if it prompts the patient to undertake breathing exercise, the best solution would be to have a device that most easily promotes the correct breathing pattern.

Surgery induces an abnormal breathing pattern characterized by a shallow, monotonous breathing without the normal periodic sighs or deep breaths involving the diaphragm (Overend et al., 2001). An effective breathing pattern to increase lung inflation should mimic sighing with a slow deep inspiration and a slow expiration maintaining a positive alveolar inflating pressure for a prolonged period (Bartlett, Gazzaniga, and Geraghty, 1973; Do Nascimento Junior et al., 2014). As the lower lung is the most common site of atelectasis, increased lower lung venti- lation should be the target for breathing exercises, and the inspiratory flow rate is an important factor deter- mining the distribution of air flow in the lung. A study of the distribution of a 133Xe bolus in normal subjects showed that an inspiratory flow of less than 300 ml/s increased the ventilation of the dependent lung while faster flows increased airway resistance and reduced ventilatory distribution (Hughes et al., 1972). Moreover, it is important to encourage full lung infla- tion and to hold that volume for several seconds, which is thought to open and stabilize the peripheral airway and alveoli using collateral ventilation in both animal and human lungs (Platell and Hall, 1997; Terry and Traystman, 2016; Terry et al., 1978; Van Allen, Lindskog, and Richter, 1931; Van Allen and Soo, 1933). Deep breathing exercise with maximum inspira- tory hold for 3–5s leads to clinically useful increases in oxygen tension (Bartlett, Gazzaniga, and Geraghty, 1973; Terry and Traystman, 2016) as well as increasing airway secretion clearance (Lapin, 2002). However, in general, patients in the acute postoperative stage are understandably reluctant to breathe deeply and inflate the lower lung and consequently, an inspiratory spi- rometer may be helpful in encouraging a beneficial breathing pattern. Deep breathing by itself has been shown to reduce atelectasis and improve oxygenation in the early stages of recovery of patients after coronary bypass surgery (Westerdahl, Lindmark, Almgren, and Tenling, 2001; Westerdahl et al., 2005, 2003).

Postoperative accumulation of secretions as a result of reduced lung volume, and a reluctance of the patient to take deep breaths and cough is a major reason for the development of postsurgical pulmonary complica- tions. This problem has been addressed previously with a pressure threshold device (BreatheMAX) in which secretion clearance is helped by the humidification and oscillation of the inspired flow as it bubbles through water (Jones, Kluayhomthong, Chaisuksant, and Khrisanapant, 2013). The sound produced by the

bubbles also provides auditory feedback to the patient, helping to maintain a steady flow of air, encouraging slow and deep inspiration, which is the key to increas- ing lung expansion (Chang, Palmer, McNaught, and Thomas, 2010).

We have investigated the use of BreatheMAX (BM) as an incentive spirometer comparing it with Triflo II (TF), an FIS device, and conventional diaphragmatic deep breathing (DB) exercise as ways of encouraging lung expansion, particularly the lower lung. Previous studies comparing FIS and VIS devices have used young subjects (Yamaguti et al., 2010), but the subjects selected for the present study were healthy older people since age is associated with an increased risk of post- surgical complications (Ji et al., 2013; Kodra, Shpata, and Ohri, 2016; Serejo et al., 2007). The objectives of the study were, therefore, to compare the pressure, flow and volume of inspired air, together with the breathing effort and accessory muscle activity, while using BM, TF and DB in healthy older subjects to assess how they might contribute to a package of measures designed to minimize postsurgical pulmonary complications.

Methods

Fourteen subjects were recruited (4 men and 10 women, age 69 ± 4 years). Subjects were all healthy with no history of cardiopulmonary diseases or smoking during the last 5 years. The study was approved by the Ethical Committee of Khon Kaen University and the subjects provided written consent before the start of any procedure. This study was a 3-way crossover trial; the three arms of the study being diaphragmatic breathing, breathing with BreatheMAX or with Triflo II with the treatment order randomized.

A sample size of 16 was calculated on the basis that the primary outcome was the difference in tidal volume between BM and TF. However, BM is a new device that has not previously been used in a study involving deep breathing exercise, and so we used values from another study of VIS and FIS (Tomich et al., 2010) that made use of a similar design and outcome measures. Of the original 16 subjects recruited, 2 dropped out and the results from the remaining 14 are reported here.

Breathing exercises

The Triflo II (Hudson RCI, USA) was used as an incentive flow-oriented device and consists of three connected plas- tic chambers each with a colored ball providing visual feedback of the flow, the target being to raise one or more balls to the top of the chamber. The manufacturer’s instructions indicate that raising one ball requires an inspiratory flow of 600 ml/s; two balls, 900 ml/s; and

2 C. UBOLSAKKA-JONES ET AL.

three balls, 1200 ml/s. Subjects were instructed to inhale deeply at a sufficient flow rate to raise 1–2 balls using abdominal or lower costal expansion and hold for 3 s at the end of inspiration, similar to (Parreira, Tomich, Britto, and Sampaio, 2005).

BreatheMAX consists of a plastic bottle with two tubes passing through the lid. The outlet tube connects to the mouth piece and the inlet tube is underwater so the depth of the water provides a threshold inspiratory load. The inspiratory air flow is humidified as it passes through the water and the bursting bubbles cause oscillations in the air flow; the structure and use of the device are described by (Jones, Kluayhomthong, Chaisuksant, and Khrisanapant, 2013). The incentive is provided by the bubble sound, which provides feedback to the patient, helping to control the pattern of breathing. Bench tests and preliminary human experiments have shown that at flow rates encoun- tered during normal breathing, a water pressure of 4–6 cm H2O generates oscillation of the air flow at 12–15 Hz, similar to that of ciliary beat frequency, and so the inspira- tory load was set at −5 cm H2O. Subjects were instructed to inspire using abdominal or lower costal expansion as far as possible and as slowly as possible while maintaining a steady stream of bubbles in the bottle, and hold the inspira- tion for further 3 s at the end of inspiration.

For the diaphragmatic breathing exercise, subjects were instructed to perform slow and deep diaphragmatic inspirations, through the mouth, to hold the maximum inspiration for 3 s and then breathe out. Subjects were in a sitting position, with a hand on the upper abdomen to palpate the abdominal movement and provide incentive feedback, as described by Tomich et al. (2007).

Procedure and protocol

After an initial interview, the subjects were familiarized with the measurements to be made and the three breathing exercises. The instructions for the subjects

and targets for breathing were similar to those used in a clinical setting. In all cases, the subjects were encour- aged to inhale deeply with a steady inspiratory flow using abdominal and lower costal expansion as much as possible, with an inspiratory to expiratory time of 1:2. With DB and BM, the inspiration was to be slow, while for TF the target was to raise the two balls. Subjects performed 3 sets of 10 breaths with 1 min rest between sets in each type of breathing to prevent possible light-headedness or dizziness as a result of hypocapnea. Subjects rested for 30 mins between each type of breathing (Figure 1) to allow sufficient time to wash out any metabolites produced by respiratory mus- cle exercise, to relieve tension in the shoulders or neck due to prolonged sitting while using the apparatus (Figure 2), and to recover a normal breathing pattern before the next type of breathing exercise. The order of the tests was randomized and balanced.

Outcome measures

Subjects breathed through a mouthpiece connected to a short tube containing airflow and pressure transducers; the tube was open to the air for the deep breathing exercise or attached to the inlet of the Triflo II or BreatheMAX, as appropriate (Figure 2). Inspiratory pressure and flow, tidal volume (Vt), inspiratory time (Ti), respiratory rate (RR), were monitored during each breathing exercise. Inspiratory mouth pressure and air flow were measured by pressure and flow transducers (TSD 104A and TSD137H; Biopac System Inc, USA) calibrated with known air pressure and gas volumes. Signals were AD converted and displayed with the BIOPAC software system (BIOPAC System, Inc., Goleta, USA.). All breathing pattern variables were analyzed from the middle 3 of the 10 breaths in each set.

Figure 1. Testing procedures. In each segment, subjects performed breathing exercise with either Deep Breathing (DB), BreatheMAX (BM), or Triflo II (TF). Vt, tidal volume; RR, respiratory rate.

PHYSIOTHERAPY THEORY AND PRACTICE 3

Ratings of perceived breathlessness (RPB) were recorded immediately after the three blocks of 10 breaths during each breathing exercise using a modified Borg scale consisting of a vertical line labelled 0–10 (Borg, 1982; Burdon et al., 1982) with verbal descrip- tors at each point; the three values were averaged to give a single value for each intervention. Heart rate (HR) and percutaneous oxygen saturation (SpO2) were measured using a bedside monitor (Nihon Kohden-Life Scope®) and recorded before and after each breathing exercise. Respiratory rate was measured using a respiration transducer (TSD 101B; Biopac System Inc, USA) and averaged over 30 s.

The use of accessory muscles was categorized as: mild, moderate, or vigorous by direct observation and palpation of the sternocleidomastoid muscle. “Mild” was used to describe a contraction where the movement of the muscle was just visible and tension just palpable; “Moderate” was a clearly defined muscle contraction; and “Vigorous” when the sternocleidomastoid and other muscles of the neck were clearly straining.

After the last breathing exercise, and with all data collected, subjects were asked to record their experience of the three breathing exercises in a short question- naire, on a 4-point scale, concerning the level of per- ceived effort or work involved in each exercise, and whether they experienced any adverse effects such as dry mouth or dizziness.

Data analysis

Data are presented as the average of nine breathing cycles, three from the middle of each of the three sets (Figure 1), except for HR and SpO2, which were recorded just before the start and immediately at the end of each exercise set. Where data were normally distributed, they are presented as mean ± SD or SEM,

as appropriate. Normally distributed data were first analyzed with repeated measures ANOVA and where significance was indicated, differences between the dif- ferent forms of breathing exercises were identified by paired Student’s t-tests with Bonferroni correction. Significance was assumed at p ≤ 0.05.

Results

Subjects were healthy older people without airflow limita- tions and their characteristics are given in Table 1. For the DB exercise, all subjects were able to breathe in slowly and hold the inspiration for 3 s. With BM, all subjects were able to breathe slowly and maintain a steady stream of bubbles in the bottle; 8 of the 14 were able to maintain the maximal inflation for 2–3 s. For the TF breathing, 3 of the 14 subjects were able to raise the two balls with every breath, the target being achieved, on average, in 70% of the breaths; only one subject was able to sustain the maximal inflation when using TF.

Inspiratory flow and pressure

Inspiratory flows and pressures differed markedly between the different breathing exercises (Figure 3).

Figure 2. Experimental set up for breathing exercises. A. Subject using Triflo II and showing the position of the flow and pressure transducers. B. Subject breathing with BreatheMAX.

Table 1. Subject characteristics. Variables Values (Mean ± SD)

Age (years) 69 ± 4 Gender (M/F) 4/10 Height (cm) 155.5 ± 5.2 Weight (kg) 62.8 ± 8.5 BMI (kg/m2) 26.0 ± 3.4 HR (bpm) 82 ± 11 SBP/DBP (mmHg) 130 ± 15/69 ± 13 RR (bpm) 18 ± 6 SVC (L) 2.4 ± 0.6 IC (L) 1.8 ± 0.4

RR: respiratory rate; SVC: slow vital capacity; IC: Inspiratory capacity. Data are given as mean ± SD.

4 C. UBOLSAKKA-JONES ET AL.

Mean flow was lowest for BM at 0.28 ± 0.1 L.s−1; for DB the mean flow was 0.43 ± 0.20 L.s−1, which did not differ significantly from the mean flow at rest (0.4 ± 0.1 L.s−1). The mean flow for TF was 0.96 ± 0.1 L.s−1, which was significantly faster than at rest, DB and BM (p < 0.001). The mean inspiratory pressures generated during inspiration with DB, BM, and TF were: 1.3 ± 0.6; 5.5 ± 1.2; and 12.8 ± 3.6 cm H2O, respectively (Figure 3). The mean inspiratory pressures for BM and TF were significantly greater than at rest, 0.9 ± 0.2 cm H2O (p < 0.001), and the TF pressure was significantly higher than BM and DB (p < 0.001).

Breathing pattern

Data related to the different patterns of breathing are shown in Table 3. The mean respiratory rate for DB and BM dropped significantly (p < 0.001) to approximately half the value at rest. When using TF, the respiratory rate was significantly greater than DB and BM (p < 0.001) and similar to the value at rest. Inspiratory time (Ti) nearly doubled during DB and more than tripled for BM compared to at rest, but was shorter for TF (BM > DB > Rest > TF, all p < 0.001). The measures of expiratory times (Te), shown in Table 3, include any breath-hold time and were longest for DB followed by BM, while both were greater than TF, which did not differ from the values at rest (DB > BM> Rest and TF, all p < 0.001). Although breath-hold time was not specifi- cally measured, Te times were consistent with subjects managing the 3 s breath-hold for DB, while for BM there was less success, and with no, or very little, breath-hold with TF.

Although the inspiratory flow was two to three times greater with TF compared with DB and BM, the Vt with TF was significantly smaller than DB and BM (Table 3. DB > TF, p = 0.033; BM >TF, p = 0.016).

However, all the breathing exercises increased Vt sig- nificantly compared to the value at rest (p < 0.001).

The Pressure Time Product (PTP) differed markedly between conditions (Table 3). When calculated from mean pressure and Ti, the values per breath were approximately four times greater than DB for TF and eight times for BM. Taking into account the breathing frequency, PTP per minute was greater for RF com- pared with BM, while both were much higher for DB

Breathing effort and accessory muscle contraction

During the breathing exercises, it was observed that the subjects made mild or moderate use of their accessory muscles, specifically the sternocleidomastoid, when breathing with DB as was the case with BM, except for three subjects who made vigorous contractions. All subjects made vigorous use of their accessory muscles of respiration when breathing with TF.

Immediately after the end of every 10 breaths, sub- jects were asked to rate their perception of breathless- ness on the RPB scale, which was generally very mild, with median values (IQR) of 0 (0–0.5) for DB, 0 (0–1.0) for BM and 1 (0.5–2) for TF. At the end of each type of the breathing excercise, trial subjects were asked to retrospectively rate their breathing effort, on a 4-point scale (Figure 4). For DB, subjects predominantly (79%) reported the exertion as none. For BM, the majority (64%) reported it to be mild, while for TF, subjects were split between reporting it as mild (36%), moderate (43%), and vigorous (21%). There were no reports of dizziness or other adverse effects of the breathing exercises.

Heart rate did not differ from resting values during any of the three respiratory exercises. Oxygen satura- tion, however, was slightly, and significantly, increased after DB (p = 0.001) and BM (p = 0.004), but was unchanged when using TF (Table 2).

Figure 3. Pressure–flow relationships. A. Inspired pressure as a function of flow. B. Tidal volume as a function of flow. Mean pressure and flow are shown for individual subjects during deep breathing (solid triangles) or using BreatheMAX (solid circles) or Triflo II (open circles).

PHYSIOTHERAPY THEORY AND PRACTICE 5

Discussion

Incentive spirometry (IS) is routinely used in post- operative respiratory care, especially following upper abdominal and thoracic surgery, with the aim of pre- venting atelectasis, the most common pulmonary com- plication, but the value of IS has been questioned (Do Nascimento Junior et al., 2014; Freitas, Soares, Cardoso, and Atallah, 2012; Jenkins et al., 1989). Part of the problem may be that the breathing pattern or lung inflation achieved with some types of IS devices may

not match the physiological requirements to properly inflate the lung and prevent atelectasis, specifically the need for deep, slow abdominal breathing. The ideal breathing pattern to achieve lung expansion is slow and deep inspiration, which allows airflow into the lower lung, and a hold for at least 3 s to increase collateral ventilation. The three breathing exercises examined here met these requirements to varying degrees (Table 3). Both DB and BM encourage slow and long inspiration, with BM being superior in this respect, probably due to the feedback provided by the bubbling sounds. However, while there was no problem with breath-hold during DB, subjects found it more difficult with BM and almost impossible with TF. Tidal volumes were significantly larger when using DB and BM compared with TF (Table 3), despite the much higher inspiratory flows (Figure 3A).

There have been a number of previous studies compar- ing VIS and FIS devices. (Parreira, Tomich, Britto, and Sampaio, 2005) report considerably shorter Ti with TF and the target set at 900 ml/s, compared with VIS measured

Figure 4. Perceptions of breathing exercise. Number of subjects rating the exercises as mild, moderate, or vigorous while diaphragmatic breathing (DB, open bars), using BreatheMAX (BM, filled bars) or Triflo II (TF, cross-hatched bar).

Table 2. Heart rate and oxygen saturation before and after breathing exercise.

DB BM TF

Pre Post Pre Post Pre Post

HR mean 67.1 64.3 64.4 62.9 65.3 65.6 SEM 3 2.5 3.1 3.1 3.1 2.9 SpO2 mean 97.9 99.1* 98.4 99.4* 98.4 98.6 SEM 0.38 0.32 0.37 0.23 0.36 0.42

Heart rate (HR) and oxygen saturation (SpO2) before (Pre) and after (Post) each of the three breathing exercises; Diaphragmatic breathing (DB), BreatheMAX (BM and Triflo (TF). Data are mean ± SEM of 14 observations. *Significantly greater than Pre measurement.

Table 3. Patterns of breathing. Rest DB BM TF

RR (brpm) 17.2 ± 1.0 6.7 ± 0.6ab 7.6 ± 0.8ab 19.5 ± 1.2 Vt (L) 0.64 ± 0.04 1.28 ± 0.08ab 1.37 ± 0.11ab 1.12 ± 0.06a

Ti (sec) 1.85 ± 0.10 3.31 ± 0.26abc 5.53 ± 0.51abc 1.16 ± 0.06a

Te (sec) 1.79 ± 0.22 6.37 ± 0.55abc 3.19 ± 0.25abc 2.06 ± 0.14 Ttot (sec) 3.69 ± 0.21 9.76 ± 0.75ab 8.77 ± 0.71ab 3.26 ± 0.19 Ti/Ttot 0.53 ± 0.05 0.34 ± 0.02 0.63 ± 0.02 0.37 ± 0.01 Peak Ins P (cmH2O) −0.58 ± 0.03 −0.83 ± 0.07

abc −8.9 ± 0.40abc −21.1 ± 1.1a

Mean Ins P (cmH2O) 0.86 ± 0.03 1.25 ± 0.16abc 5.54 ± 0.31abc 12.82 ± 0.96a

PTP/breath (cmH2O.sec) −1.57 ± 0.10 −3.71 ± 0.29 abc −30.32 ± 3.19abc −15.47 ± 1.70a

PTP/min (cmH2O.sec.min −1) −27 ± 2 −25 ± 3abc −207 ± 11abc −280 ± 21a

RR: respiratory rate (breaths per minute); Vt: tidal volume; Ti: inspiratory time; Te: expiratory time (includes any breath-hold); Insp P: inspiratory pressure; PTP: pressure time product. Superscript a, differs significantly from Rest; b, differs significantly from TF; c, significant difference between DB and BM. Data are mean ± SEM.

6 C. UBOLSAKKA-JONES ET AL.

by the Coach and Voldyne devices, although Vt was much the same, implying their young subjects, unlike the present older subjects, were able to generate sufficiently high flows to compensate for the shorter Ti in every breath. Interestingly, the authors report increased SpO2 when using the two VIS and TF. In the present study, DB and BM were also found to increase SpO2, but this did not occur with TF, probably because our older subjects, unlike younger subjects, were not able to fully inflate when using TF. Paisani et al. (2013) reported longer Ti when using VIS and likewise, Tomich et al. (2007) found Ti to be shortest when subjects used TF, longer with DB, and longest for VIS breathing. Tidal volume was largest for VIS, and lowest for TF, although the differences were not statistically different, which was probably because the relatively young subjects (18–44 years) were able to sustain the high pressure and flow for longer, as with the Parreira, Tomich, Britto, and Sampaio (2005) study. The only previous study that has included older subjects who may be more representative of high-risk patient groups was that of (Lunardi et al., 2014). As with other studies, they found Ti to be longest with VIS, although the actual Ti times were surprisingly short at around 2 s compared to 5–7 s reported here for DB and BM.

The longer Ti with VIS devices compared with TF indicates slower inspiration, although our present study appears to be the only one where actual flow rates have been measured. The importance of slow inspiration lies largely in the fact that a flow of less than 300 ml/s is required to increase ventilation of the dependent lung (Hughes et al., 1972), the area most susceptible to atelectasis. Fast inspiration involves the activation of accessory muscles of inspiration, which tend to increase expansion of the upper costal region rather than the lower regions (Roussos et al., 1977). These authors reported that an inspiration from FRC with enhanced abdominal expansion increases distribution of inspired gas to the dependent lung zones while, when intercostal and accessory muscles are used, there is an increased gas distribution to nondependent lung regions. Costal expansion is therefore associated with greater ventila- tion of the upper lung and less to the base of the lungs (Tomich et al., 2010). Consistent with this, (Parreira, Tomich, Britto, and Sampaio, 2005) found that there was greater abdominal expansion when breathing with VIS devices compared with TF and (Paisani et al., 2013) reported longer Ti when using VIS, which was asso- ciated with greater abdominal expansion compared with FIS.

Tomich et al. (2007) showed increased activity of the sternocleidomastoid muscle during FIS breathing, as did (Lunardi et al., 2014). This is consistent with the observa- tions of the present study of vigorous use of the accessory

muscles when using TF. The mean pressure when using the TF Triflo II was −12 cm H2O, but the peak pressure generated at the start of the inspiration was often as high as −20 cm H2O (Table 3) as the subjects attempted to reach the target of raising two balls in the device and was achieved by a substantial involvement of the accessory respiratory muscles, which would be expected to expand the upper costal region rather than the lower regions of the lung (Roussos et al., 1977). Weindler and Kiefer (2001) have documented the fact that flow-oriented IS devices increase the work of breathing as a consequence of the high pres- sures required. If healthy subjects have difficulty achieving full lung inflation when using TF, it is unlikely that patients, one or two days post-surgery, will be able, or willing, to generate the high inspiratory pressures and be even less successful at fully inflating their lungs than our healthy subjects.

BM breathing also involves generating an inspiratory pressure, in this case about −6 cm H2O, which subjects generally described as requiring mild effort. However, lung inflation was marginally, although not significantly, higher than with DB and with significantly longer Ti time (Table 3), which may favor expansion of the lower lung. The load of −5 cm H2O was chosen to provide a signifi- cant increase in transpulmonary pressure and oscillate the airflow at 12–15 Hz, which may resonance with the ciliary beating frequency and help secretion clearance (George et al., 1985). Our healthy older subjects had no trouble with this load, but patients may find it a challenge, in which case the load could be reduced according to the ability of the patient. It is possible that TF could also be used more effectively if the instructions were altered so that instead of trying to raise two balls, the patient attempted just the first ball. This would reduce the air flow to 600 rather than 900 ml/s, although even this might not be optimal for inflation of the lower lung; elevating the first ball half way up the tube might be a more suitable target when using TF. It is notable that in the study of Parreira, Tomich, Britto, and Sampaio (2005), the inspira- tory flow when using TF was reported as around 500 ml/s, suggesting that the target was to raise one ball, although this was not explicitly stated.

DB imposes virtually no resistance to inspiration yet Ti was significantly shorter than with BM, despite subjects having the same instructions to breathe slowly and deeply. The difference probably arises from the nature of the feed- back. With DB, the subjects obtain some feedback from their hands on the abdomen, but this is not very effective because toward the end of the breath, there is little move- ment of the abdomen and therefore little sensory feedback via the hand. In contrast, the subjects have very direct feedback with BM as they have to maintain the sound of air bubbling through the water.

PHYSIOTHERAPY THEORY AND PRACTICE 7

The difficulty experienced by the subjects when using TF may have occurred for two reasons. The first is a result of the relatively high peak and mean pressures required (Table 3), entailing stronger muscle contractions and invol- ving the accessory muscles, which is experienced as an increased effort. Differences in the inspiratory pressure, Ti and breathing frequency, leading to a higher PTP per minute with TF (Table 3), may increase the metabolic demand and also contribute to the greater sense of effort. These factors probably account for why the subjects were unable, or unwilling, to hold their breath at the end of inspiration when using TF.

The inability of our subjects to breath-hold at the end of inspiration when using TF was surprising but it is not clear whether this was unusual. Subjects were not asked to breath-hold in the study by Parreira, Tomich, Britto, and Sampaio (2005), but in the Lunardi et al. (2014) study, subjects were instructed to breath-hold for 5 s, although whether or not they achieved this is not recorded. However, if the total breathing cycle time given in Table 2 of that paper included breath-hold time, it would appear that the older subjects were not breath-holding when using TF since the Ti and estimated Te were similar to those in the present study.

One notable difference between the present study and of those by Parreira, Tomich, Britto, and Sampaio (2005) and Lunardi et al. (2014) is in the breathing frequency when using TF. The lower frequency of the Parreira study is partly explained by the slower inspiratory flow rate of around 500 ml/sec compared to 960 ml/sec in the present study. In the Lunardi study, the breathing frequency of the older subjects using TF was around 14 brpm compared with 19 in the present study, but the resting frequency for their subjects was also lower, so possibly this represents different natural breathing rates of the two subject groups.

Limitations of this study

This study was performed for healthy older people; there- fore, the findings may differ from postsurgical patients with pain, although we anticipate that this would increase the difference between TF and the other two breathing techni- ques. Future studies of either healthy subjects or patients would benefit from a larger sample size and the assessment of accessory muscle work and breathing effort were sub- jective rather than quantified using EMG measurement or Borg RPE. Since breath-hold may be an important part of breathing exercises, it would be best to specifically monitor this rather than including the breath-hold as part of Te, as done here. The primary objective of the breathing exercises was to inflate the lower lung and future work should aim to asses this either directly, by measurements of regional ven- tilation,orbythecarefulassessmentofchestandabdominal

expansion. Future studies require randomized trials of the clinical efficacy of the different types of breathing exercises.

In summary, even when the subjects were given similar instructions to breathe deeply, we have found that TF does not inflate the lung as fully as either DB or BM, and it is likely that the high flow rates of TF mitigate against the expansion of the lower lung. TF did not increase oxygena- tion, while itwasimprovedwith DB and BM. Moreover, the high pressures involved with TF and the possible greater metabolic cost may lead to an increased perception of effort in meeting the target flow, and is unlikely to encourage or motivate patients who need to undertake frequent breath- ing exercises to reduce the risk of developing atelectasis. For well-motivated patients who understand the need for slow and deep breathing, DB could be effective but the feedback provided by a hand on the abdomen is not very sensitive and does not regulate the inspiratory flow rate. TF could be effective ifthetargetflow ratewas considerably reduced and slower flows may be achieved with volume-orientated devices. BM provides clear feedback of flow and inspiratory time to both patient and physiotherapist and has the advan- tage of a humidified and oscillating inspired air flow, which couldassistwith secretionclearance,another factor that will help reduce the incidence of pulmonary complications. Further clinical study of the efficacy of different breathing techniques is needed in patients with thoracoabdominal surgery.

Declaration of interest

Dr Ubolsakka-Jones developed BreatheMAX, used in this study, and is the academic advisor to the company manufacturing the device.

References

Bartlett RH, Gazzaniga AB, Geraghty TR 1973 Respiratory maneuvers to prevent postoperative pulmonary complica- tions. A critical review. Jama 224: 1017–1021.

Borg GA 1982 Psychophysical bases of perceived exertion. Medicine and Science in Sports and Exercise 14: 377–381.

Branson RD 2013 The scientific basis for postoperative respiratory care. Respiratory Care 58: 1974–1984.

Burdon JG, Juniper EF, Killian KJ, Hargreave FE, Campbell EJ 1982 The perception of breathlessness in asthma. American Review of Respiratory Disease 126: 825–828.

Cassidy MR, Rosenkranz P, McCabe K, Rosen JE, McAneny D 2013 I COUGH: reducing postoperative pulmonary complica- tions with a multidisciplinary patient care program. Journal of American Medical Association Surgery 148: 740–745.

Chang AT, Palmer KR, McNaught J, Thomas PJ 2010 Inspiratory flow rate, not type of incentive spirometry device, influences chest wall motion in healthy individuals. Physiotherapy Theory and Practice 26: 385–392.

Do Nascimento Junior P, Módolo NS, Andrade S, Guimarães MM, Braz LG, El DR 2014 Incentive spirometry for prevention of postoperative pulmonary complications in upper abdominal

8 C. UBOLSAKKA-JONES ET AL.

surgery. Cochrane Database of Systematic Reviews 2: CD006058.

Freitas ER, Soares BG, Cardoso JR, Atallah ÁN 2012 Incentive spirometry for preventing pulmonary complica- tions after coronary artery bypass graft. Cochrane Database of Systematic Reviews 9: CD004466.

George RJ, Johnson MA, Pavia D, Agnew JE, Clarke SW, Geddes DM 1985 Increase in mucociliary clearance in normal man induced by oral high frequency oscillation. Thorax 40: 433– 437.

Hughes JM, Grant BJ, Greene RE, Iliff LD, Milic-Emili J 1972 Inspiratory flow rate and ventilation distribution in nor- mal subjects and in patients with simple chronic bronchi- tis. Clinical Science 43: 583–595.

Jenkins SC, Soutar SA, Loukota JM, Johnson LC, Moxham J 1989 Physiotherapy after coronary artery surgery: are breathing exercises necessary? Thorax 44: 634–639.

Ji Q, Mei Y, Wang X, Feng J, Cai J, Ding W 2013 Risk factors for pulmonary complications following cardiac surgery with car- diopulmonary bypass. International Journal of Medical Sciences 10: 1578–1583.

Johnson RG, Arozullah AM, Neumayer L, Henderson WG, Hosokawa P, Khuri SF 2007 Multivariable predictors of postoperative respiratory failure after general and vascular surgery: results from the patient safety in surgery study. Journal of the American College of Surgeons 204: 1188– 1198.

Jones CU, Kluayhomthong S, Chaisuksant S, Khrisanapant W 2013 Breathing exercise using a new breathing device increases airway secretion clearance in mechanically ventilated patients. Heart and Lung: Journal of Critical Care 42: 177–182.

Kelkar KV 2015 Post-operative pulmonary complications after non-cardiothoracic surgery. Indian Journal of Anaesthesia 59: 599–605.

Kodra N, Shpata V, Ohri I 2016 Risk Factors for Postoperative Pulmonary Complications after Abdominal Surgery. Macedonian Journal of Medical Sciences 4: 259–263.

Lapin CD 2002 Airway physiology, autogenic drainage, and active cycle of breathing. Respiratory Care 47: 778–785.

Lawrence VA, Cornell JE, Smetana GW 2006 Strategies to reduce postoperative pulmonary complications after noncardiothor- acic surgery: systematic review for the American College of Physicians. Annals of Internal Medicine 144: 596–608.

Lunardi AC, Porras DC, Barbosa RC, Paisani DM, Marques Da Silva CC, Tanaka C, Carvalho CR 2014 Effect of volume-oriented versus flow-oriented incentive spirometry on chest wall volumes, inspiratory muscle activity, and thoracoabdominal synchrony in the elderly. Respiratory Care 59: 420–426.

Miskovic A, Lumb AB 2017 Postoperative pulmonary complica- tions. British Journal of Anaesthesia 118: 317–334.

Overend TJ, Anderson CM, Lucy SD, Bhatia C, Jonsson BI, Timmermans C 2001 The effect of incentive spirometry on postoperative pulmonary complications: a systematic review. Chest 120: 971–978.

Paisani DDM, Lunardi AC, Silva CC, Porras DC, Tanaka C, Carvalho CRF 2013 Volume rather than flow incentive spirometry is effective in improving chest wall expansion and abdominal displacement using optoelectronic plethys- mography. Respiratory Care 58: 1360–1366.

Parreira VF, Tomich GM, Britto RR, Sampaio RF 2005 Assessment of tidal volume and thoracoabdominal motion using volume and flow-oriented incentive spirometers in healthy subjects. Brazilian Journal of Medical and Biological Research 38: 1105–1112.

Platell C, Hall JC 1997 Atelectasis after abdominal surgery. Journal of the American College of Surgeons 185: 584–592.

Poelaert J, Szegedi L, Blot S 2013 Prevention of postoperative pulmonary problems starts intraoperatively. In: Annual Updates in Intensive Care and Emergency Medicine p. 539– 552. New York, Springer Verlag, Vincent, Jean–Louis (Ed.).

Roussos CS, Fixley M, Genest J, Cosio M, Kelly S, Martin RR, Engel LA 1977 Voluntary factors influencing the distribution of inspired gas. American Review of Respiratory Disease 116: 457–467.

Serejo LG, Da Silva-Júnior FP, Bastos JP, De Bruin GS, Mota RM, De Bruin PF 2007 Risk factors for pulmonary complications after emergency abdominal surgery. Respiratory Medicine 101: 808–813.

Terry PB, Traystman RJ 2016 The clinical significance of collateral ventilation. Annals of the American Thoracic Society 13: 2251– 2257.

Terry PB, Traystman RJ, Newball HH, Batra G, Menkes HA 1978 Collateral ventilation in man. New England Journal of Medicine 298: 10–15.

Tomich GM, França DC, Diniz MT, Britto RR, Sampaio RF, Parreira VF 2010 Effects of breathing exercises on breath- ing pattern and thoracoabdominal motion after gastro- plasty. Jornal Brasileiro De Pneumologia 36: 197–204.

Tomich GM, França DC, Diório AC, Britto RR, Sampaio RF, Parreira VF 2007 Breathing pattern, thoracoabdominal motion and muscular activity during three breathing exercises. Brazilian Journal of Medical and Biological Research 40: 1409– 1417.

Van Allen CM, Lindskog GE, Richter HG 1931 Collateral respiration. Transfer of air collaterally between pulmonary lobules. Journal of Clinical Investigation 10: 559–590.

Van Allen CM, Soo YC 1933 Collateral respiration: spontaneous reinflation of an atelectatic pulmonary lobule by collateral respiration. Journal of Clinical Investigation 12: 171–179.

Weindler J, Kiefer RT 2001 The efficacy of postoperative incentive spirometry is influenced by the device-specific imposed work of breathing. Chest 119: 1858–1864.

Westerdahl E, Lindmark B, Almgren SO, Tenling A 2001 Chest physiotherapy after coronary artery bypass graft surgery–a comparison of three different deep breathing techniques. Journal of Rehabilitation Medicine 33: 79–84.

Westerdahl E, Lindmark B, Eriksson T, Friberg O, Hedenstierna G, Tenling A 2005 Deep-breathing exercises reduce atelectasis and improve pulmonary function after coronary artery bypass surgery. Chest 128: 3482–3488.

Westerdahl E, Lindmark B, Eriksson T, Hedenstierna G, Tenling A 2003 The immediate effects of deep breathing exercises on atelectasis and oxygenation after cardiac sur- gery. Scandinavian Cardiovascular Journal 37: 363–367.

Yamaguti WP, Sakamoto ET, Panazzolo D, Peixoto CDC, Cerri GG, Albuquerque AL 2010 Diaphragmatic mobility in healthy subjects during incentive spirometry with a flow-oriented device and with a volume-oriented device. Jornal Brasileiro De Pneumologia 36: 738–745.

PHYSIOTHERAPY THEORY AND PRACTICE 9