Comparison of Inspiratory Work of Breathing in T-Piece Breathing, PSV, and Pleural Pressure Support Ventilation (PPSV)

Comparison of Inspiratory Work of Breathing in T-Piece Breathing, PS'" and Pleural Pressure Support Ventilation (PPSV)* Toshimichi Takahashi, M.D.; ]un Takezawa, M.D., F.C.C.R; Tomomasa l(jmura, M.D.; l(jmitoshi Nishiwaki, M.D.; and Yasuhiro Shimada, M.D., F.C.C.R

We have compared the inspiratory work of breathing during T-piece breathing, pressure support ventilation (PSV), and pleural pressure support ventilation (PPSV) by using a lung model with variable compliance and resistance, under simulated spontaneous breathing. Our lung model consists of two spring-loaded bellows, representing the lung and diaphragm, placed in an airtight container. Inspiration begins with the withdrawal of air from the diaphragm bellows by a time-cycled jet-8ow-creating Venturi mechanism. Expiration occurs by opening the diaphragm bellows to the atmosphere. Work or breathing (WOB) is calculated by plotting the pressure-volume curve, with pressure corresponding to intrabox pressure and volume corresponding to the tidal volume; PPSV is a new mode of mechanical ventilatory support accomplished by setting the ventilator (Servo 900C) into the PSV mode with a level of 0 em HsO, using the pleural pressure as the input and target signal.

ressure support ventilation (PSV) was designed to P synchronize the patient's effort through How-

servomechanisms by monitoring and targeting the

PaW;l.2 PSV has an advantage in improving patientventilator synchrony3-5 and extends its clinical application to ventilatory assistance in patients with pulmonary patholo~6 The use of Ppl, instead of Pa~ as a target pressure allows us to accomplish a new mode of mechanical ventilation called pleural pressure support ventilation (PPSV);7 PPSV is accomplished by putting the ventilator (Servo OOOC) into a PSV mode at either a pressure support level of0 cm H 20 (PPSVo) or greater than 0 cm H 20. Although PPSVo is targeting Ppl in attempting to maintain the constant Ppl of0 cm H 20, a positive Paw is actually added to the airway to overcome the resistance and compliance of the lung during inspiration, resulting in a reduction of the lung work to zero, but the chest wall work will remain. When PPSV of greater than 0 cm H 20 is used (which is accomplished by putting the ventilator into the PSV mode at the pressure support level of greater than 0 *From the Intensive Care Unit, Nagoya University Hospital, Nagoya, Japan. Manuscript received September 1; revision accepted February 4. Reprint requests: Dr. Takahashi, leU, Nagoya University Hospital, 65 Tsuromai-cho, Showa-ku, Nagoya 466, japan

1030

The PPSV muimally reduces WOB under any circumstances. The PSV sufficiendy reduced WOB only in the oormallung and the lung with low compliance; however, a pressure supporting time is diminished in the lung with low compliance. The serious limitations of PSV remain in its application to the lung with high resistance. It is concluded that PPSV is closer to the actual patient's signal and has a potential advantage in reducing WOB in the lung with low compliance or high resistance (or both). The lung with 80w limitation is still a challenging issue for mechanical ventilatory assistance. (Ched 1991; 100:1030-34)

= =

=

lung compliance; Cw chest wall compliance; lIE ratio inspiration-expiration ratio; PPSV pleural pressure support ventilation; PPSV. = PPSV of 0 em 0.0; PSV =:= pressure support ventilation; SIN ratio = signal to noise ratio; V airway

~

=

Bow

=

cm H 20 with targeting the Ppl), an additional positive pressure is added to the Paw and will result in a reduction of the chest wall work as well. The characteristic comparison of T-piece breathing, PS~ and PPSV is shown in Table 1. In comparison with PS~ PPSV may have the advantages of (1) less delay or distortion (or both) of the patient's effort, and (2) a target pressure directly related to the patient's ventilatory work load (ie, PPSVo reduces a patient's lung work to zero). In this stud~ we used a lung model which simulated spontaneous breathing with variable compliance and airway resistance and compared work of breathing (WOB) in these three modes (ie, T-piece breathing, PS~ and PPSV) in an attempt to elucidate the advantages of PPSV over T-piece breathing and PS~ Our investigation also attempted to address the Table l-Compariaora of Concept and CharacterVtica of TPiece Breathing, PS~ and Authors' De8igned PPSV*

Data

Location ofTarget Pressure

Target Pressure

Consequent Inspiratory Paw

Consequent Inspiratory ppl

T-piece PSV PPSV

Airway Airway Pleura

Insp=exp Insp>exp Insp=exp

Insp= exp Insp>exp Insp>exp

Insp
*Insp, Inspiratory; and exp, expiratory. Inspiratory 'Nark of Breathing (Takahashi at eJ)

X-y RECORDER, STRIP CHART RECORDER

SERVO BOOC Ppl

Paw (P,I> .anitlril. (PPSY)

TillE-CYCLED JET- FLOW GEIERATOR

used to create a Venturi effect so that the spontaneous T-piece

breathing, which was accomplished by disconnecting the ventilator from the lung model, was simulated with a VT of 700 ml. Next, a ven~or (Siemens-Elema Servo 900C) was connected to the lung bellows to deliver PPS~ The PPSVo was delivered by putting the ventilator into a PSV mode using a zero-pressure support level in which ppl was monitored and targeted. Sensitivity was 1 em 8.0, and the driving pressure of the ventilator was 15 cm 8.0 during PPS~ During PPSVo, the driving pressure of the jet-Row generator was adjusted to obtain a VT of 700 mi. The same driving pressure of the jet-flow generator during PPSVo was used during PSV; however, the level of PSV was adjusted to return the VT to 700 mi. Sensitivity was 1 em R.O, and the driving pressure of the ventilator was 60 em HsO during PS~ Subsequentl~ a pressure support level was raised above and below the level to obtain the VT of 700 ml. Finally, to evaluate a further reduction in WO~, the level ofpressure support used during PPSV was increased from 0 to 2 and 4 em H.O. The mechanical property of the lung model was changed as follows: normal (CL , Q.12 Uem 8.0; R, 5 em 8.0IlJs); high resistance (eL , 0.12 Uem HsO; R, 20 em HsOllJs); low compliance (CL , 0.04 Ucm HIO; R, 5 em H.OllJs); and high resistance with low compliance (CL , 0.04 Ucm H.O; R, 20 em RIOllJs). In all of the experimental procedure, the respiratory rate was 20 breaths per minute, and Cw was 0.12 Uem H.O. RESULTS

FIGURE 1. Schematic representation of layout of equipment; Paw

was monitored and target~ during PS~ and ppl was monitored

and targeted during PPS~ V, Airway flow; and ppl, pressure inside airtight box.

limitations of PSV in reducing WOB in the diseased lung. MATERIALS AND METHODS

Apparatw Figure 1 shows the scheme ofour lung model. It consisted of two bellows placed in a 2-L airtight plastic container. The upper bellows represented the lung, and the lower bellows represented the chest wall, including the diaphragm. Both bellows were suspended by the springs. The compliance of the chest wall (Cw) was adjusted by changing the elasticity of the spring and set at 0.12 Uem B.O. The lung compliance (C.) was also varied by changing the elasticity of the spring and set at either 0.12 or 0.04 Uem 8 10. Airway resistance was created by interposing a resistor with a resistance of either 5 or 20 em ",OllJs between the flowmeter and the lung model. A jetflow Venturi effect created a negative pressure inside the lower bellows for making inspiration through an approximately 20- L reservoir box as a capacitance. During expiration, the jet flow was terminated, and the lower bellows was open to the atmosphere, allowing the bellows to return to the origiDallevel. The following variables were measured: VT; airway flow Paw; pressure inside the box (PPI); and WOB. The Paw and ppl were measured by a pressure transducer. The VT was measured by a hot-wire ftowmeter (Minato ATD 1(5) by integrating the flow signal. These signals were processed into a microcomputer (Minato RM-300) and recorded on a multichannel strip-chart recorder (Sanei Omnicorder). A PPI-VT curv~ was obtained by the microcomputer for calculating WOR with electrical integration.

M;

Protocol The experiment was performed with the following sequence: Fint, we adjusted the driving pressure of the jet-Sow generator

A typical tracing of V and Paw versus time and of the ppl-VT re1ations~p during T-piece breathing, PPS~ and PSV in the lung model ofnormal compliance (both C L and C w are 0.12 Ucm HIO) and normal airway resistance (5 cm HIOlUs) is shown in Figure 2. In T-piece breathing, the Paw remained unchanged at 0 mm Hg, and the ppl Huctuation was 3 mm Hg (from - 4 to - 1 mm Hg). Both Paw ~d ppl Huctuations were smaller in PPSVo than those in PSV of 11 cm HIO (PSV11) at a matched VT. The same tendency was equally found in all of the following experimental settings when ~mpliance or resistance or both were varied. When the pressure support level during PPSV was increased to 2 and 4 cm H 20 (PPSVI and PPSVJ, Paw was substantially increased, resulting in a higher ppl and a larger VT. This higher PPSV may further reduce the chest wall work. This observation was found equally in all of the following experimental settings with changing the mechanical property of the lung model. In PS~ a pressure support level of 11 cm HIO was needed to obtain a VT of 700 ml. When the pressure support level was changed to less or more than 11 cm HIO during PS~ the corresponding changes in VT were observed. The Ppl-VT curves during T-piece breathing, PPSVo, and PSV11 with a matched VT are also shown in Figure 2. Inspiratory WOB at a matched VT of 700 ml was 1,736 g-cm in Tpiece breathing, 783 g-cm in PSV11 , and 192 g-cm in PPSVo• When resistance was increased to 20 em HIOlUs with a normal compliance (Fig 2), ppl Huctuation in Tpiece breathing was - 1 to - 8 mm Hg. A pressure support level of 12 cm HIO was required to obtain a VT of 700 ml during PS~ Incidentall~ a significant CHEST I 100 I 4 I OCTOBER. 1881

1031

)1111

fLJL.

.~ Ii

C••,lialce 0.12Uc.H21 Resistalce 5 c••21/Us

f II~

11---

!

111 ~

_

fl~

'-PIa I.l•

C••plialce 0.12 UC.H28 Resista.ce 20c.H20/Us

.; -111

2'~ 11

.. f

I

I

"U ,.1II1) J ... " 5453 ml

'·PlECE

JII:1 U

m.. ,••

"",

112

-II

) :.

.J;:

1I1~ I

I

JV\

'"", ',I"

11:~~

r

r'~ T mYl~ -51

I

I

"",

m, -II

JII:1 U

I

+11 ",CCill.)

m".,.lII

m r L2U

mooT

-51

"".. "t.

PSt,.,••

,

-II

!

'Ilil"

I

III] JlJL,

t HI "'CCill.)

ppsy,. ,.lU

I

i 111 11

!

I

!_

1"~ I

f .. ~

~l.

!

I

-II

.;;" -III

21 11 I

HI "'CCill.)

h1L-

»

PSt".,.lll

'.C.)

T

I

U L L III J" UII 1153

PSt.

:;: -111

j I:~ -1LJL

+11 "'CClIIII')

!lJl-

+11 p,!CClIII.)

"".. ,.m

IIU ':CII":: J" Il"

11I

tJ +11 p'!CCIIlh.)

111]

I

t ::1~

" I W ,.clln:111 i

f

~

!

:.: -111

j 1:~ jLJL "LL "cllnJ III " 1111

-II

PS'"

"LU III " 1111 2144

,

~

+51 "'CCill.)

-II

PSt,.. I • •

"".. '.IJI

I

.! II: ~ JL.tL

+11 "'CClIII.)

m.. "'II

I-':~~ ll~ I

ULLJ" UII PStIi

PPSY.

, J -II

'1l·):2

"w '.C.) ':1 "UII

"W ,.clln III" 1111 511 PSt••

~ +11 ",CCill.)

-II

~ I

HI "IlCllll.)

-51

I

+51 "'lCllll.)

FIGURE 2. Two columns at left show analog tracings of V and Paw vs time, and ppl-VT relationship curves during T-pieee breathing, PS~ and PPSV in lung model of compliance of0.12 Ucm "sO and resistance of 5 em "sOllJs. Two columns at right show those in lung model of compliance of 0.12 Uem "10 and resistance of 20 em "20IlJS. WOB, Work of breathing (g-cm).

Paw oscillation was observed during PSV of7, 12, and 20 cm H 20. During PS~ inspiratory time of PSV was prolonged due to a long time constant, resulting in positive pressure sustaining at the start of expiration. Inspiratory WOD in T-piece breathing, PSV12, and PPSVo was 5,453 g-cm, 2,244 g-cm, and 467 g-cm, respectivel~

As shown in Figure 3, in the case of reduced compliance and normal airway resistance, Ppl 8uctuations in T-piece breathing were - 2 to - 16 mm Hg. A pressure support level of 24 cm H 20 was required to obtain the matched VT of 700 mI. During PS~ inspiratory demand flow terminated earlier than the end of inspiration of the lung model, resulting in imposing the inspiratory WOB at any pressure support levels examined. The WOD during T-piece breathing, PSV24 , and PPSVo was 8,373 g-cm, 2,124 g-cm, and 646 g-cm, respectively. 1032

When resistance was increased and compliance was reduced (Fig 3), Ppl 8uctuation was from - 3 to - 19 mm Hg during T-piece breathing. A pressure support level of22 cm H 20 was required to obtain the matched VT of.700 mI. The Paw oscillation was also observed during PS~ During PS~ a pressure supporting time returns to that in the lung model ofnormal compliance and resistance. This is because both lung models had a similar time constant. Inspiratory WOB during Tpiece breathing, PSV22 , and PPSVo was 11,786 g-cm, 1,644 g-cm, and 1,202 g-cm, respectivel~ Figures 2 and 3 explained that both PPSV and PSV reduced inspiratory work by supplying a demand flow to assist contracting muscles (diaphragm bellows). The PPSVo versus PSV with the same inspiratory effort and VT, however, had several important differences. First, the triggering time, which was defined as an interval from the start of a negative deflection of Inspiratory Wort< of Breathing (Takahashi et 81)

i ~

1111 11.rvL.

HIlU I,lIS

31] 21

..

f I

11

Co.,liaICI 0.04Uc.12D Resistaici 5 C.121/Us

_ _~

I

JLn..

!

111] I 4: -111

T·PlU I,Ja

COllpliance 0.04 UC.H20 Resistalce 20 C.H20/US

21~

f

..

I

11 I

_.---'_

-!

111~ I

![IV

',1.) ..,

UW" ''I' 1113

T·'llCl

-51

•

'I

+'1 "'I".) mY.. 1,121

~II:~~ 31] Ii 1121 I

mt,

"w ',111I)::"''I' ~ -~.

t

t

•

.; -111

1lJl

i

t -'I

j:HM

"'"

Lf

J-Vv

-u

1I1~

! -II: i

I

I~W" "1.) ..2124,15"

H' ".1....)

•

mY..

',NI

:HM ',I.)w:

-u

•

~11:~ ~ f

I

"w ,,1·):2 " lUI

-'I

-~ 111~ I .; -10e

1t -II

•

" W,

mY!

.1

m.. 1,114

"] fUl 21 11 I

111 ~ I ';-111 _

!

i

21

~

•

~

PPSl..

i !

1:~

I

nI' PPSl.

r f'~ r -'I

•

I

-~.

,'1'

+u "'1"')

1I:~ ~ m.. ~1lI

"l M

-II' 21 II I

',1.):

'51'

"UJ"

PSY"

,~

+'1 "'1. . .)

',1.) ...

t J

H' ".1....)

!

+11 "'I".)

1144

"'n I,.

•

" W

'5

',1.) . .,1111

-U

f " -u

~

H' ".1....)

',I.):

"w" nI'

+u ",I".)

•

It/v

II

"LLJ"

mY.

",.

I ::~N~

"

\

m.. I , .

21 11 I

!

~

.g -111J -

mY!

~

-II'

~ 1:~

H' ".1....) PSYJ.. ,,111

"w ',1.)-:,1'1' "

..! II: ~ i "j hJL

mY,

\, •

mY,.I,JU

21

I

"W "1.):''1' "

H' "'1".)

h h V V

11:]

m.

m.. I,"J

H' "II".)

t -II

L

•

+11 "'I".)

3. Two columns at left show analog tracings of V and Paw vs time, and ppl-VT relationship curves during T-piece breathing, PS~ and PPSV in lung model of compliance of 0.04 Ucm H 20 and resistance of 5 em H 20/I/s. Two columns at right show those in lung model of compliance of 0.04 Ucm HIO and resistance of 20 em H 20/IJs. WOB, Work of breathing (gecm). FIGURE

Ppl to its negative peak during PSV and PPS~ was always shorter during PPSV (200 ms) compared with that during PS~ except that in a low compliance lung (100 ms). The triggering time in PSV was longest in the high-resistance lung (400 ms) and shortest in the low-compliance lung. When the compliance was reduced and resistance was increased, the triggering time was 150 ms in PS~ A variable triggering time during PSV may be caused by a difference in the time constant of the lung and may impose an additional inspiratory work. On the other hand, the triggering time in PPSV was approximately 200 ms in any lung conditions. The relatively constant triggering time in PPSV is associated with the location of the sensing catheter, which is closer to the contracting muscles (diaphragm bellows), and may lead to a similar reduction in triggering work under any lung conditions.

Secondly, PPSVo has a slower inspiratory flow due to a lower target pressure and a lower driving pressure of the ventilator. This lower flow allows the inspiratory time to be prolonged and the inspiratory airway pressure to be lowered due to the fact that the 25 percent peak flow cutoff occurred at a much lower How rate. In PS~ 25 percent cutoff occurs at a higher How rate, resulting in earlier termination of the demand flo~ even though the lung is still in the inspiratory phase. This phenomenon is particularly true in the lung with low C L • These findings are in contrast to PPS~ in which the lIE ratio remains the same during any lung conditions. This is again due to the location of target pressure sensing, which is close to the contracting muscles (diaphragm bellows), and, therefore, can minimize distortion of inspiratory efforts caused by ahway resistance and compliance. CHEST I 100 I 4 I OClOBER, 1991

1033

DISCUSSION

In the current stud~ we calculated WOB by integrating ppl multiplied by the VT by the following formula: work = pressure change x volume change. 8.9 Therefore, the work calculated in our study includes the lung work and imposed work. In the normal lung, 45 percent of WOB measured during T-piece breathing was attained by 11 cm H 20 of PS~ corresponding mostly to the work needed to trigger the ventilator. This is in good agreement with the work done by Marini et al,10.11 in which they compared WOB between spontaneous breathing and assist-control mode ventilation. On the other hand, PPSV required only 20 percent ofWOB needed during T-piece breathing. Although an esophageal balloon can be used to measure Ppl in clinical situations, it cannot steadily trigger a ventilator because of its low SIN ratio. To trigger the ventilator more efficiently (with less work), the feasibility of other signals, such as the diaphragmatic EMG, should be pursued. 12-1. Therefore, PPSV can be used as a reference ventilatory mode in evaluating a WOB reduction accomplished by a variety of other ventilatory modes; however, Ppl Huctuation monitored by the esophageal balloon may reHect the efficacy of ventilatory support, especially in diseased lungs. It is therefore recommended that ppl should be periodically monitored to confirm that the pressure assist ventilation being delivered is optimal. In PSV the Ppl showed the second negative deHection at the end-inspiratory phase (Fig 2 and 3). Several explanations can be ~ade for this phenomenon. First, because lung compliance is relatively low (especially at 0.04 Vcm H20), the ventilator terminates the demand flow even though the lung is still in the inspiratory phase (the demand How rate reached to the terminating criteria of 25 percent of initial inspiratory How earlier than the end of actual inspiration). If this occurs, additional inspiratory work is added at the end of inspiration. Secondly, even though a pressure support tenninates at the end of inspiration, the expiratory Ppl curve returns to the original expiratory ppl curve drawn duringT-piece breathing. Ifa pressure support duration exceeds the inspiratory phase, it ma~ on the other hand, prolong the inspiratory time, resulting in an increase in an auto-PEEP level in the high-resistance lung. 15.16 A small increment of PPSV levels resulted in a large increase in VT, even in the diseased lung; however, this effect was not accompanied by the large Ppl Huctuations observed during PSV; however, it is unknown whether this overshooting of ppl due to a higher PPSV level affects the cardiac performance. Both PPSV above 0 cm H 20 and PSV with a higher support level progressively reduces a patient's work. 1034

Thus, chest wall work requirements of the patient become reduced with these higher levels. In summary, we compared the inspiratory WOB in T-piece breathing, PS~ and PPSV using a lung model with variable compliance and airway resistance. The PPSV can maximally reduce WOB in any conditions oflung status. Reduction in WOB attained by PSV was almost satisfactory in the lung with normal compliance and can be sufficiently reduced in the low-compliance lung, if terminating How rate can be reduced to zero; however, a serious limitation remains in the reduction in WOB in the lung with high airway resistance. Mec~ca1 ventilatory assistance in the lung with How limitation is still a challenging issue, even for PS~

REFERENCES

1 Takezawa J, KimuraT, SanoT, Nabshima K, Komatsu T, Shimada Y. Comparison of the work of breathing between pressure support ventilation and SIM~ Crit Care Med 1989; 17:S33 2 MacIntyre NR, Nishimura M, Usuda Y, Tokioka H, Takezawa J, Shimada Y. The Nagoya conference on system design and patient interactions during pressure support ventilation. Chest 1990;

1J1: 1463-66

3 MacIntyre NR. Respiratory function during pressure support ventilation. Chest 1986; 89:677-83 4 MacIntyre NR. Pressure support ventilation: effect on ventilatory reflexes and ventilatory-muscle workloads. Hespir Care 1987; 32:447-57 5 Kacmarek RM. The role of pressure support ventilation in reducing work of breathing. Respir Care 1988; 33:99-120 6 Brochard L, Pluskwa F, Lemaire F. Improved efficacy of spontaneous breathing with inspiratory pressure support. Am Rev Respir Dis 1987; 136:411-15 7 Takezawa J, 18bhashi T, IGmura T, Sano T, Nabshima K, Shimada Y. Pleural pressure support ventilation: a new mode of mecbanical ventilation to support spontaneous breathing. Crit Care Med 1989; 17:833 8 Braun NMl: Faulkner J, Hughes RL, Roussos C t Sahgal ~ When should respiratory muscle be exercised? Chest 1983; 84:76-84 9 Leith DE, Brardley M. Ventilatory muscle strength and endurance training. J Appl Physioll976; 41:508-16 10 Marini D, Rodriguez R, Lamb ~ The inspiratory workload of patient initiated mechanical ventilation. Am Rev Respir Dis 1986; 134:902-09 11 Marini JJ, Capps J, Culver B. The inspiratory work of breathing during assisted mechanical ventilation. Chest 1985; 87:612-18 12 Petit JM, Millic-Emilli J, Delhez L. Role of the diaphragm in breathing in conscious normal man: an electromyographic stud~ J Appl Physioll960; 15:1101-06 13 Broebard L, Half A, Lorino H, Lemaire F. Inspiratory pressure support prevents diaphragmatic fatigue during weaning from mecbanical ventilation. Am Rev Respir Dis 1989; 139:513-21 14 Gurevitch MJ, Gelmont D. Importance of trigger sensitivity and ventilator response delay in advanced chronic obstructive pulmonary disease with respiratory failure. Crit Care Med 1989; 17:354-59 15 Smith TC, Marini D.Impact of PEEP on lung mechanics and work of breathing in severe airflow obstruction. J Appl Physiol

1988; 65:1488-99

16 Pepe PE, Marini JJ. Occult positive end-expiratory pressure in mechanically ventilated patients with airflow obstruction. Am Rev Respir Dis 1982; 126:166-70