BENCH TESTING OF THE CPU-1 VENTILATOR

Br. J. Anaesth. (1986), 58, 653-662

BENCH TESTING OF THE CPU-1 VENTILATOR J. F. NUNN AND D. J. R. LYLE

DESCRIPTION OF THE APPARATUS

The CPU-1 ventilator comprises a pneumatic section which occupies the lower part, and a microprocessor section which occupies the upper part of the apparatus (fig. 1). A simplified diagram of the pneumatic section is shown in figure 2. Artificial ventilation is accomplished by the flow regulator which delivers gas to the patient at a pre-set flow rate, either for a pre-set time or until a pre-set pressure is reached. In parallel to this circuit is the demand valve, which permits spontaneous breathing at any time. Thus Intermittent Mandatory Ventilation is an integral feature of the CPU-1. Ventilation, either spontaneous or artificial, is monitored by a hot wire anemometer distal to the expiratory valve. For technical reasons there is a continuous flow of 2 litre min"1 through the demand valve. Operation of the flow regulator and the expiratory valve is controlled by the microprocessor. The operator can select the mode of the ventilator and the appropriate controls are then illuminated (table I). However, within certain modes, the microprocessor assumes responsibility for adjusting the artificial ventilation in response to the patient's spontaneous ventilation and other factors. The operator has only limited control over the pattern of response of the CPU-1, which in many cases is complex and not intuitively obvious to staff. J. F. NUNN, MJ>., PH.D., F.R.C.S., F.F-A.R.C.S., F.F.A.R.A.CS.

(HON.), F.F.A.R.C.S.I. (HON.), Division of Anaesthesia, Clinical Research Centre, Watford Road, Harrow, Middlesex HA1 3UJ. D. J. R. LYLE,M.B.,B.S.,P.F.A.R.C.S., Department of

Anaesthetics, Crawley Hospital, Sussex.

spontaneous breathing of the patient, and provides a wide variety of operating modes. These include spontaneous breathing with or without continuous positive pressure, volumecycled and pressure-cycled artificial ventilation, with or without positive end-expiratory pressure or synchronization. In addition, there is a facility for ensuring a mandatory minute volume. The ventilator is controlled by a microprocessor with sophisticated decision-taking functions based on input from an expiratory hot-wire anemometer. Intended and measured ventilatory variables are displayed separately. The anemometer has satisfactory accuracy and pressure control is adequate. The synchronization permits adaptation of the ventilator to spontaneous respiration according to a wide range of harmonics. In any mode, artificial ventilation is initiated if spontaneous respiration is inadequate. The mandatory minute volume mode is based on a complicated program of a cautious nature which reacts instantly to inadeqate spontaneous breathing, but is slow to discontinue artificial ventilation.

A bank of nine small option selector switches on the rear of the pneumatic section of the apparatus control certain important functions of the ventilator. The operator should be aware of their settings at all times, although they are not visible from the front of the apparatus. Functions set by these switches include the availability of the Mandatory Minute volume mode, the monitoring of inadequate tidal volume, the manner of calculation of the patient's respiratory frequency and the characteristics of the "sigh". Various modes of operation may be selected by

Downloaded from http://bja.oxfordjournals.org/ at University of Iowa Libraries/Serials Acquisitions on June 7, 2015

The authors had tie opportunity of assessing the original version (V4) of the CPU-1 ventilator in SUMMARY 1982. Following discussions with the designer, An artificial patient capable of spontaneous and M. Maillot, a number of minor modifications were artificial ventilation has been used for bench incorporated in the version currently available testing of the Ohmeda CPU-1. This ventilator has (V5), which is the subject of this report. an extensive capacity for interaction with the

654

BRITISH JOURNAL OF ANAESTHESIA

Fresh gas

Preset MMV (Inspjratory and expiratory times; Inspiratory flow rate K

DISPLAY Expired Imposed Minute volume Minute volume Tidal volume Tidal volume Frequency Frequency

transducer

Patient FIG. 2. Simplified diagram of the pneumatic circuit.

Downloaded from http://bja.oxfordjournals.org/ at University of Iowa Libraries/Serials Acquisitions on June 7, 2015

FIG. 1. The CPU-1 ventilator.

655

CPU-1 VENTILATOR TABLE I. Controls actwe in different model Inspiration time

Inspiratory pause

Expiration time

Inspiratory effort

(Inspiratory) Flow rate

Peep

Sigh

means of the mode selector switch located at the bottom left of the pneumatic section. They are as follows: Constant positive airway pressure (CPAP) This is the normal mode for spontaneous breathing and provides infinitely variable CPAP from zero to about 30 cm H,O. Artificial ventilation is commenced automatically in the event of inadequate spontaneous respiration. Inadequacy is denned by up to three entirely separate criteria. First, the apparatus will always respond to a sustained absence of expiratory flow (defined as less than 6 litre min"1 instantaneous flow) which continues for more than about 10 s. The subsequent response includes the institution of artificial ventilation and is described in detail below. Second, the CPU-1 will always provide a visual and audible alarm when the minute volume decreases below the setting of the small Minimum Minute Volume knob located in the centre of the microprocessor section. The third system is switched in and out by one of the option selector switches. When the tidal volume decreases below a pre-set value there are audible and visual alarms, together with the automatic institution of artificial ventilation. Clearly this system (controlled by option selector No. 9) should remain in the " ON " position. Controlled ventilation This mode provides time-cycled constant flow ventilation. Inspiratoryflowrate, inspiratory time,

inspiratory pause and expiratory time can all be pre-set by the operator. However, inspiration is also terminated if the airway pressure reaches the level pre-set by rotating the bezel of the anaeroid gauge. Alarms operate as described above under the CPAP mode. Synchronized ventilation This mode differs from the previous mode in the prolongation of expiration by 1 s, during which time an inspiratory effort by the patient will trigger the next inflation, thereby enabling the ventilator to synchronize itself to the patient's spontaneous respiration within certain limits. In the absence of triggering, the prolongation of expiration decreases the respiratory frequency and, therefore, the minute volume. Mandatory minute volume (MMV) MMV is selected as a secondary option of the previous mode by pressing the yellow button behind the mode selector switch. In this mode the level of artificial ventilation is automatically adjusted to ensure that the total minute volume is always equal to or greater than a pre-set value selected by the "Minimum Minute Volume" control located in the microprocessor section. It is necessary to distinguish two separate settings of minute volume on the CPU-1 in this mode. First, there is the Minimum Minute Volume, set by the small knob in the centre of the display panel of the microprocessor section. In the long term this defines the Mandatory Minute

Downloaded from http://bja.oxfordjournals.org/ at University of Iowa Libraries/Serials Acquisitions on June 7, 2015

"CPAP" (Continuous Positive Airways pressure) "CONTROLLED" (Time-cycled constant flow ventilation) "SYNCHRONISED" (Controlled + trigger) "MMV" (Mandatory Minute Volume) "PRES CYCL CONT" (Pressure Cycled) "PRES CYCL SYNC" (above with trigger)

656

BRITISH JOURNAL OF ANAESTHESIA Minute Volume setting appears as a very fast flickering display (about 10 Hz) while being set and at intervals thereafter. Inadequate tidal volume according to the criteria described below is shown with four "decimal points" in the display. In addition, a wide range of numerical error codes may appear in the tidal volume display. A separate red display, "APNEA", appears under the various circumstances listed in this paper.

Pressure-cycled controlled ventilation In this mode, inspiratory time is limited solely by the airway pressure attaining a value pre-set by rotating the bezel of the pressure gauge. Expiration remains time-cycled and, together with the inspiratory flow rate, is under the control of the operator.

METHODS OF EVALUATION

Pressure-cycled synchronized This mode again differs from the previous mode in the prolongation of expiratory time by 1 s, during which time an inflation may be triggered by a spontaneous inspiration. Displays Airway pressure is displayed by a conventional anaeroid manometer set in the pneumatic section of the CPU-1. The remaining displays are light-emitting diodes in the microprocessor section. On the left (in yellow) are the pre-set values for minute volume, ventilator tidal volume, ventilator frequency and inspiratory: expiratory ratio. These values are calculated from the settings of the time and flow rate controls and indicate what is intended and not what the patient actually receives. Reduction of frequency by selection of the synchronized mode is automatically calculated and displayed to indicate the minimum frequency if there is no triggering. There is no yellow display in the pressure-cycled mode. Actual ventilation is measured by the expiratory hot wire anemometer. There is an in-built correction for the continuous flow of 2 litre min"1 which passes the demand valve. Corrected values are then displayed in red on the right of the microprocessor section as minute volume, tidal volume and frequency. This display includes both spontaneous and artificial ventilation, each expiration being displayed separately. A minute volume less than the selected Minimum Minute Volume setting appears as a flashing display (about 2 Hz). The Minimum

The accuracy of the hot wire anemometer was assessed by ducting expired gas from the anemometer directly to a 10-litre water-sealed spirometer which filled stepwise. The mean of at least six breaths measured by the spirometer was then compared with the concurrent display on the monitor. The anemometer was calibrated according to the manufacturer's instructions immediately before assessment. Airway pressures were measured with a strain gauge pressure transducer calibrated against a column of water. To determine the pattern of interaction with a patient, the CPU-1 was used in conjunction with the model lung described by Lyle and his colleagues (1984). This device consists of a spring-loaded bellows with a compliance which can be varied between 25 and 70 ml/cm H2O and variable airway resistance. In addition to being ventilated artificially, the model lung can breathe spontaneously with a tidal volume infinitely variable up to 2300 ml and a frequency range of 4.6-84 b.p.m., although maximal tidal volume can only be attained at frequencies up to 13 b.p.m. There is provision for a spontaneous inspiration to be inhibited by lung inflation, and separate electrical outputs are provided for tidal excursion and "diaphragmatic contraction". Thus there is no difficulty in distinguishing between spontaneous and artificial breaths. The model patient was also used for supplying "expired" gas for assessment of the hot wire anemometer. RESULTS

The hot wire anemometer Values transduced from the hot wire anemometer were generally within 10% of the value measured by the spirometer (fig. 3). There was, however, a tendency to over-read at high volumes and under-read at low volumes. The spirometer should

Downloaded from http://bja.oxfordjournals.org/ at University of Iowa Libraries/Serials Acquisitions on June 7, 2015

Volume. Second, there is what the instruction book describes as the "Safety Setting" which comprises the settings of inspiratory and expiratory times and inspiratory flow rate as used in the mode from which MMV was entered. The CPU-1 provides artificial ventilation at the full value of these settings when the MMV mode is first entered and also when spontaneous ventilation decreases to less than the selected Minimum Minute Volume setting. Details are given below.

CPU-1 VENTILATOR

657 1000n y-x-100 800-

600-

200-

0

Simulated spontaneous breathing o Artificial ventilation (time-cycled) • Artificial ventilation (pressure-cycled)

200

400

600

800

1000

Volume measured by water spirometer (ml) FIG. 3. Expired tidal volumes measured by the hot wire anemometer compared with values obtained from a water-sealed spirometer. The line labelled y = x is the line of identity without correction for the flow of 2 litre min"1 through the demand valve. The labelled line y = x —100 includes the correction which is built into the CPU-1 display.

have been affected by the 2 litre min"1 continuous flow through the demand valve (follow the line labelled y = x—100), but did not seem to be so. Neither is it apparent in table III below (p. 659). Pressure values in CPAP mode With the ventilator in CPAP mode and the PEEP control (positive end-expiratory pressure) set at zero, airway pressure was + 3 m HSO during expiration and — 2 cm H8O during inspiration (relative to atmosphere) (fig. 4, B). This differential was approximately maintained during increasing application of PEEP up to the maximum point of the analogue scale which corresponded to + 22 cm H,O during expiration and +17 cm H,O during inspiration (fig. 4, E). The maximal PEEP of + 30 cm H,O could be obtained by moving the control beyond the end of the scale as far as the stop. Response to inadequate ventilation in the CPAP mode (1) Apnoea. When spontaneous breathing was abruptly interrupted, the ventilator waited 12 s and then gave a single inflation (fig. 5). A second inflation occurred after another 12 s and, thereafter, regular ventilation took place according to the setting of the ventilator controls. The alarm

sounded after about six to 10 artificial breaths. When it was silenced, the entire apnoea sequence was re-entered. It was impossible to maintain artificial ventilation in the CPAP mode without the alarm sounding. In this mode artificial ventilation was at a frequency of 13 b.p.m., the expiratory duration control being inoperative. (2) Minute volume less than Minimum Minute Volume setting. There was an immediate response when the actual minute volume (in the red display) became less than the setting of the Minimum Minute Volume. The response comprised the audible warning and a flashing minute volume display. "APNEA" was not displayed, and there was no automatic institution of artificial ventilation. (3) Inadequate tidal volume. This modality was selected with option switch No. 9. Switch No. 4 was then used to select either 1/4 or 3/8 of the specified tidal volume as "inadequate". The specified tidal volume appears in the yellow display on the left of the microprocessor unit and is determined as the product of the inspiratory flow rate and duration of inspiration by adjustment of the relevant controls. After about seven "inadequate" breaths, the response was triggered

Downloaded from http://bja.oxfordjournals.org/ at University of Iowa Libraries/Serials Acquisitions on June 7, 2015

400-

BRITISH JOURNAL OF ANAESTHESIA

658

1500 -i

1000-

25

500

o CD

a

I

a .S Q

Alarm

500-

ID Z3

0

1

2 3 4 Elapsed time (min)

5

6

7

8

FIG. 5. Response to apnoea in the CPAP mode. The model lung breathing spontaneously during the first 1 min (tidal volume 625 ml). The CPU-1 was pre-set to deliver a tidal volume of 500 ml in the event of apnoea. Arrows indicate the times when the alarm was silenced.

Downloaded from http://bja.oxfordjournals.org/ at University of Iowa Libraries/Serials Acquisitions on June 7, 2015

1min FIG. 4. Volume and pressure traces during simulated spontaneous breathing in the CPAP mode. A = Ventilator off, but with gas supply connected; B = CPAP mode, PEEP, set to zero; C, D , E = CPAP mode, progressive settings of PEEP controls on the analogue scale.

659

CPU-1 VENTILATOR TABLE II. Activation of alarm by inadequate tidal volume in CPAP mode (ml)

selected as "inadequate"

Denned inadequate tidal volume

820 610 410 205 800 610 400 310 255

3/8 3/8 3/8 3/8 1/4 1/4 1/4 1/4 1/4

307 229 153 77 200 152 100 77 63

295-300 210-220 145-155 60-S5 190-205 150-165 95-100 65-80 60-75

Fraction

within the range of tidal volumes shown in table II, the tidal volume being read from the red display. The response comprised the appearance of four "decimal points" in the tidal volume display, the signal "APNEA", the audible warning and the initiation of artificial ventilation. Controlled ventilation mode

There was no difficulty in effecting a wide range of different patterns of ventilation, although only with constant inspiratory flow rate. Table III shows close agreement between pre-set tidal volumes (yellow display), monitored tidal volumes (red display) and actual tidal volumes (measured by spirometry). There was no difficulty in ventilating the artificial patient with a minute volume of 18 litre min"1 when the compliance was reduced to 25 ml/cm H4O and the airway resistance increased to 50 cm H,O at a flow rate of TABLE III. Tidal volumes in time-phased ventilation mode (ml). Values obtained at a frequency of 21 b.pjn. by varying inspiratory flow rate. Compliance: 75 ml/an HtO

Pre-set

Displayed on monitor

Measured with spiromcter

155 270 330 450 490 570 680 790 890 940

175 275 340 450 530 590 790 870 1000 1050

205 290 350 440 510 560 730 820 950 980

Synchronized mode

When controlled breathing was superimposed on spontaneous breathing at a different frequency, confused asynchronous respiratory patterns were obtained which represented the patient "fighting the ventilator" (fig. 6). Over a wide range of different frequencies, entry to the synchronized mode resulted in the ventilator synchronizing itself at various harmonics of the patient's spontaneous respiratory frequency. Figure 6 shows entry to an alternating pattern of spontaneous and triggered breaths, but it was also possible to obtainratiosof 1:3,2:3 and even more complex harmonics. The expiratory level reverts to the functional residual capacity in figure 6 as expiration is able to proceed to completion. The trigger responds to an attempted inspiration in less than 1 s. Mandatory Minute Volume (MMV) mode There is an infinite range of combinations of settings of Minimum Minute Volume, initial or safety settings and the patient's spontaneous minute volume. We report here some examples which illustrate the most important interactions and their time courses. Figure 7 shows entry into MMV mode when the artificial patient's spontaneous minute volume was 6 litre min"1 and the Minimum Minute Volume had been set at 5 litre min"1. Artificial ventilation was immediately started at the initial (or "Safety") setting, substantially increasing the

Downloaded from http://bja.oxfordjournals.org/ at University of Iowa Libraries/Serials Acquisitions on June 7, 2015

Specified tidal volume

Tidal volume range within which alarm was activated

30 litre min 1. Spontaneous breaths could be taken at any time during controlled ventilation. Tidal volume was progressively diminished by the introduction of a leak at the connection between the gas circuit and the model lung, thus simulating a leak at the endotracheal tube. At an intended tidal volume of 810 ml, there was no alarm until the monitored expiratory tidal volume decreased to less than 65 ml. The response was a red tidal volume display of 0000, the signal "APNEA" and the audible warning. This was independent of the setting of selector switch No. 9, which is inoperative in this mode. This also occurred in the synchronized mode. When the monitored expired minute volume was less than the setting of the Minimum Minute Volume knob, the monitored expired minute volume display flashed, but there was no other visible or audible warning. This applied to all the modes providing artificial ventilation.

BRITISH JOURNAL OF ANAESTHESIA

660

750-

C

o * 500$

I 250-

g 3

0J Elapsed time (min)

FIG. 6. Use of the synchronized mode. The first part of the trace (A) shows lung volume during spontaneous breathing (tidal volume 250 ml at 26 b.p.m.). In the centre section (B) artificial ventilation was superimposed (tidal volume 350 ml at 20 b.pjn.). In the third section ( Q the synchronized mode was selected and alternate spontaneous breaths were augmented by artificial ventilation at 13 b.p.m.

I—I Patient's spontaneous breathing EUIPPV

Apnoea

- MMV setting

Time in MMV mode (min) FIG. 7. Entry into MMV mode when the patient's spontaneous minute volume is already in excess of the Minimum Minute Volume setting. The frequency of artificial ventilation is progressively reduced to 0.4 b.p.m. by the 16th minute.

total minute volume. According to the program in the microprocessor, the frequency of artificial ventilation was then decreased progressively over 16 min to a minimal value of 0.4 b.p.m. A simulated apnoea at that point initiated a procedure similar in most respects to that described above for apnoea in the CPAP mode (fig. 5). The only difference was that, in the MMV mode, silencing

the alarm did not cause reversion to the start of the apnoea procedure and artificial ventilation could continue without the alarm sounding. Artificial ventilation started at the initial (or "Safety") setting and gradually decreased itself to the selected Minimum Minute Volume value by progressive reduction of the respiratory frequency with a time scale similar to that shown in figure 7.

Downloaded from http://bja.oxfordjournals.org/ at University of Iowa Libraries/Serials Acquisitions on June 7, 2015

§

CPU-1 VENTILATOR I

661 I Patient's spontaneous breathing IPPV

MMV setting

When the situation shown at the 16th minute in figure 7 had been reached, progressive reductions in either tidal volume or frequency of the simulated spontaneous breathing, had no effect until the spontaneous minute volume decreased below the MMV setting (fig. 8). When this occurred, it was interpreted as though the patient was apnoeic and artificial ventilation was immediately resumed at the initial (or "Safety") setting. Artificial ventilation was then only gradually reduced until it reached the level of the MMV setting. Pressure-cycled modes

Nothing untoward or unexpected was found in the operation of the CPU-1 in these modes. Inspiratory flow rate was infinitely variable up to 120 litre min"1 and the limiting pressure could be set at any level up to 80 cm H2O. Duration of expiration was variable between 0.5 and 30 s. In a simulated paediatric situation, it was possible to attain a respiratory frequency of 90 b.p.m. and a tidal volume down to 60 ml. Warnings appeared for apnoea (e.g. disconnection) and a minute volume less that the Minimum Minute Volume setting gave a flashing display. There was, however, no response to inadequate tidal volume as the result of a partial leak. DISCUSSION

The CPU-1 is a sophisticated ventilator which is capable of providing a wide range of modes of

operation. It will take over certain aspects of the ventilatory management of the patient and is able to undertake a limited range of decisions in response to input from the expiratory flow transducer. However, its operation is not intuitively obvious and the instruction manual does not explain all aspects of its function. A crucial feature of the CPU-1 is the expiratory flow transducer and its performance seems to be adequate for its function (fig. 3). Under-reading at low minute volumes and over-reading at high minute volume exaggerates the departure from normality and is no bad thing. The pressure swing of 5 cm HSO in the CPAP mode (fig. 4) is typical of the results obtainable from demand valve systems. CPAP with truly constant pressure throughout the respiratory cycle is difficult to obtain unless one resorts to a weighted bellows with compensated PEEP valve such as the system described by Hewlett, Platt and Terry (1977), in which the pressure swing does not exceed 1 cm H,O. It is a remarkable safety feature of the CPU-1 that, in the CPAP mode, there should be three independent alarm mechanisms triggered, respectively, by a long expiratory pause, inadequate minute volume or inadequate tidal volume. However, it must be stressed that only the first is an obligatory feature. The second only operates if the Minimum Minute Volume is set and it provides automatic artificial ventilation only in the MMV mode. The third requires deliberate setting of two option switches and the setting of the target

Downloaded from http://bja.oxfordjournals.org/ at University of Iowa Libraries/Serials Acquisitions on June 7, 2015

Time (min) FIG. 8. The effect of gradual reduction of spontaneous minute volume below the Minimum Minute Volume setting in the MMV mode.

662

Terry (1977). The CPU-1 is excessively cautious in deciding that the patient's breathing is adequate. It is not obvious why artificial ventilation should be added to spontaneous breathing which is already above the specified Minimum Minute Volume. Overall, the CPU-1 appears to be a soundly constructed, sophisticated ventilator with elaborate control mechanisms. The advantages of these systems must be set against the considerable effort which will be required in instructing staff in its use. A period of adaptation will also be required for staff to attune themselves to the decision making capacity of the CPU-1, a development which is, nevertheless, commonplace in other walks of life. In conclusion, it may be noted that it is exceptionally difficult to determine the mode of operation of a complex interactive ventilator either in clinical use or by using it with a simple artificial lung. The development of the new model lung has greatly simplified our task and we believe that it may also have a role in the training of staff. REFERENCES Hewlett, A . M . , Plan, A. S., and Terry, V. G. (1977). Mandatory minute volume. A new concept in weaning from mechanical ventilation. Anaesthesia, 32, 163. Lyle, D. J. R., Nunn, J. F., Hawes, D. W. C , Dickins, J., Tate, M., and Baker, J. A. (1984). A model lung with the capacity for simulated spontaneous breathing. Br. J. Anaesth., 56, 645.

Downloaded from http://bja.oxfordjournals.org/ at University of Iowa Libraries/Serials Acquisitions on June 7, 2015

tidal volume. Nevertheless, these three independent mechanisms should confer a high measure of safety. A dangerous situation might develop if staff came to rely upon them and for some reason omitted to make the required settings. It should be noted that the program cannot distinguish between apnoea and a respiratory frequency of less than 6 b.p.m. Operation in the time-cycled constant flow mode calls for little comment. It is a flexible but powerful ventilator which can achieve high minute volumes in the face of very high values of respiratory impedance. It is not obvious why inspiration should be pressure-limited as well as time-cycled, since this option is available in the pressure-cycled mode. In the synchronized mode there was an impressive capacity to synchronize the ventilator to the spontaneous breathing of the patient and restore expiration to the functional residual capacity. Quite complicated harmonics could be obtained and there was almost invariably a regularization of the breathing pattern, as shown in figure 6. In the Mandatory Minute Volume mode, the CPU-1 departs radically from the original concept of maintaining the minute volume constant from minute to minute, whatever the level of the patient's spontaneous breathing. It was by no means easy to elucidate the pattern of response of the ventilator in this mode, which is far more complicated than with the original pneumatic circuit described by Hewlett, Platt and

BRITISH JOURNAL OF ANAESTHESIA