Thorax 1996;51:756-761

Techniques in mechanical ventilation: principlesand practice

J M Shneerson

The principles and practice ofmanaging mech-anical ventilation are often daunting to chestphysicians. It is easy to see why this should be.Firstly, the physician's experience of ventilatorsis usually in intensive care units where theresponsibility rests largely, but often to a looselydefined degree, with anaesthetists. The com-plex and rapidly changing medical problems ofpatients require frequent adjustments to theventilator settings. In addition, there is a widerange of ventilators and the details of theirperformance characteristics are often un-available to the user. The classification of vent-ilators is becoming more complex and moreunsatisfactory as their versatility increases, andthe nomenclature of the various modes ofvent-ilation is inconsistent and confusing.' Many ofthese have been inadequately assessed and theirclinical indications are uncertain. Lastly, thereis an increasing range of interfaces (such asnasal and face masks and new designs oftracheostomy tubes) between the ventilator andthe patient to choose from and with which fewdoctors are familiar.2

Patients with both extrapulmonary restrictivedisorders and chronic airflow obstruction are,however, being treated with ventilators in in-creasing numbers on medical wards and intheir homes, as well as in intensive care units.Decisions about how and when to initiate vent-ilation, how to wean the patient from the ventil-ator, and how to provide long term ventilatorysupport are having to be faced increasinglyfrequently. This review addresses the principlesand some of the practical issues of this type oftreatment.

Respiratory Supportand Sleep Centre,Papworth Hospital,Papworth Everard,Cambridge CB3 8RE,UKJ M Shneerson

Techniques of ventilation and theirlimitationsThe conventional descriptions of ventilatorsand ventilatory methods are complex and in-consistent. The names used by manufacturersfor the method of support can be misleadingand it is more important to understand what theventilator is actually doing. Ventilators whichgenerate a predetermined flow rate, which isusually either constant or a half sine, ramp, or

reverse ramp wave form, and which have a

fixed inspiratory time will deliver a predictabletidal volume. This volume preset ventilationcontrasts with pressure preset ventilation inwhich a predetermined pressure waveform isdelivered. The tidal volume will depend onthe impedance to inflation and the patient'srespiratory effort, but the ventilator can com-

pensate for leaks. Some degree of leak com-

pensation is also possible with volume presetventilators by providing a high tidal volume

which allows for the leak and setting a pressurelimit during inspiration as a safety measure incase the leak decreases.The ventilator may be time cycled from in-

spiration to expiration but is said to be triggeredif a change in pressure or flow due to thepatient's effort alters the phase of respiration.The patient's influence on ventilation cantherefore range from the extreme of controlledventilation where the tidal volume, frequency,inspiratory and expiratory times are fixed, topressure preset ventilation with triggering ofboth inspiration and expiration, in which thepatient can modify all of these variables. Syn-chronised intermittent mandatory ventilation(SIMV) combines volume preset breaths whichcan be triggered, with spontaneous breaths withor without a preset pressure. Other less com-monly used techniques such as airway pressurerelease ventilation, volume or pressure con-trolled inverse ratio ventilation, proportionalassisted ventilation and high frequency vent-ilation have been developed but are not inwidespread use3 and will not be consideredfurther here.Once the principles underlying the ventil-

atory methods have been established, it isimportant to ascertain the details of the per-formance characteristics of the ventilator itself.The manufacturers' manuals often give littleexplicit information and there is a remarkablepaucity of data in the published literature abouthow closely different ventilators approximateto the ideal properties that would be expectedfrom their brief technical descriptions. Thisvoid frequently leads to practical difficultiesand unexpected complications in ventilatormanagement. Ideally, the ventilator will ac-curately deliver the preset flow or pressure waveform in the face of varying patient mechanicsand changing leaks and it should be able toprovide sufficient flow for patients requiringhigh inspiratory flow rates.4 The inspiratoryresistance of the ventilator and circuit shouldbe low ifthe patient is triggering breaths and thetrigger delay should be short. The expiratoryresistance of the circuit should be low5 and theextent of rebreathing within the circuit, whichcan vary considerably,6 should be minimal.

Principles of gas exchange duringventilationThe primary aim of mechanical ventilation isto improve alveolar gas exchange. The threecomponents of the system - the ventilator to-gether with its inspiratory and expiratory cir-cuits, the patient, and the interface betweenthe two - are in series and all influence gas

756

on February 25, 2021 by guest. P

rotected by copyright.http://thorax.bm

j.com/

Thorax: first published as 10.1136/thx.51.7.756 on 1 July 1996. D

ownloaded from

Techniques in mechanical ventilation: principles and practice

exchange. The minute ventilation is the prod-uct of the respiratory frequency and the tidalvolume, but there are complexities even inthis apparently simple relationship. Altering thefrequency, for instance, changes inspiratoryand expiratory times which affect the mean

inspiratory flow rate in volume preset vent-ilators, the time for equilibration of alveolargases and, especially in the presence of airflowobstruction, the end expiratory volume.The tidal volume leaving the ventilator dur-

ing inspiration has three destinations: (1) leaks;(2) dead space ventilation; and (3) alveolarventilation.

LEAKSThe gas from the ventilator follows the path oflowest impedance which may be into the lungs,especially ifan inspiratory effort is synchronisedwith the ventilator's inspiratory phase, or itmay be lost through leaks in the system. Thesemay be within the ventilator or its circuit, atthe interface with or within the patient. Aninitial search for disconnections in the ventil-ator system should be made, but more com-

monly leaks are related to the interface withthe patient. Some air usually leaks around a

cuffed endotracheal or tracheostomy tube butthe leak is greater with an uncuffed tube. Thisis usually fairly constant and can be reducedby selecting a tube which is large relative tothe tracheal lumen. The leak does, however,vary according to the pressure generated inthe trachea by the ventilator and the patient'sefforts and the control ofupper airway patency,particularly at the laryngeal level.78 Leaksaround both face and nasal mask systems and,with the latter, through the mouth are often a

problem. They may be associated with a raisedpressure in the upper airway when this ob-structs or when the patient and ventilator are

uncoordinated, or with a low upper airwaypressure when the muscles controlling thepalate and lips fail to provide an airtight seal.This seal varies according to the degree ofsynchronisation between the ventilator and thepatient, any underlying neuromuscular disease,the stage of sleep, and probably with the in-dividual's intrinsic coordination of the upper

airway muscles. Leaks may also occur withinthe patient, for instance, through a broncho-pleural fistula or by air entering the oeso-

phagus rather than the trachea. Volume presetventilators with a high flow capacity and pres-

sure limiting may cope adequately with leaksbut, in general, they are better compensatedfor by pressure preset ventilators.

DEAD SPACE VENTILATIONThe dead space of each ventilator circuit variesand is dependent on the length and diameter ofthe connecting tubing. The gas under positivepressure stretches the circuit by 2-3 ml/cm H20inspiratory pressure (the compressible volume).This is equivalent to an increase in dead space

since this volume never reaches the patient.The second component of the dead space is

within the interface of the system with the

patient. The dead space of a nasotracheal ororotracheal tube is greater than that of atracheostomy tube and, similarly, the deadspace of a face mask is greater than that of anasal mask or mouth piece. The larger volumeinterfaces require more flow from the patientto trigger the ventilator because the pressurefall within them is inversely related to theirvolume, although this effect may be counter-balanced by the higher resistance of narrowerinterfaces, such as a small tracheostomy tube,which increases the effort needed to trigger theventilator.

Within the patient, the anatomical deadspace may be partially bypassed by a tracheo-stomy but not by non-invasive techniquessuch as nasal ventilation and, in general, thephysiological dead space is increased ratherthan decreased during mechanical ventilation,particularly if the patient becomes hyper-inflated.

ALVEOLAR VENTILATIONThe distribution of the remaining part of thetidal volume among the alveoli is determinedby factors such as the inspiratory time and theinspiratory pressure and flow pattern whichresult from the interaction of the ventilatorand the patient's respiratory efforts, and thecompliance and resistance of the alveoli. Inlung diseases these are heterogeneous and anincrease in the tidal volume may only over-distend the alveoli with the shortest timeconstants without any increase in gas exchange.Recruitment of alveoli with longer time con-stants may be achieved by prolonging the in-spiratory time or adding an end inspiratorypause.

Aims of ventilationVentilation is usually only a part of the overalltreatment and needs to be coordinated withthe other components of the management plan.Clearly defined goals are essential wheneverpatients are ventilated, although it is rare forall of these to be achieved simultaneously.Compromises are inevitable and the prioritiesof treatment and the risks of the ventilatorsettings have to be carefully assessed.The following questions should be addressed

whenever ventilation is being considered.

IS NORMALISATION OF THE ARTERIAL BLOODGASES THE AIM?In general the answer to this question is yes,and only in exceptional circumstances shouldabnormal blood gas tensions be accepted.Patients with acute on chronic respiratorydisorders are often managed with the aim ofmaintaining a degree of hypercapnia on theassumption that this is their normal state andthat lowering the Pco2 will lead to apnoeaand difficulties in withdrawing ventilatory as-sistance. This assumption should not be madeunless the premorbid blood gas tensions areavailable and, in any case, the concentration ofbicarbonate in the cerebrospinal fluid rapidly

757

on February 25, 2021 by guest. P

rotected by copyright.http://thorax.bm

j.com/

Thorax: first published as 10.1136/thx.51.7.756 on 1 July 1996. D

ownloaded from

Shneerson

equilibrates with a lowered arterial Pco2 so thatafter 24-48 hours the sensitivity to changes inPco2 increases. The Pco2 can usually be re-duced satisfactorily by ventilation althoughsupplemental oxygen may be required to main-tain normal Po2.9

Occasionally, patients with severely ab-normal resistance or compliance cannot beventilated adequately, and the peak inspiratorypressure required to provide a satisfactory tidalvolume may rise to potentially dangerous levels(40-50 cm H20). The risk of barotrauma andthe adverse effects on cardiac function may besufficient to warrant a reduction in tidal volumeor frequency, or both, to levels which allow thePco2 to rise in a planned fashion (controlledhypoventilation or permissive hypercapnia).- 1 I

Conversely, hyperventilation is occasionally re-quired for the specific indication of reducingintracranial pressure when a Pco2 of about4 kPa should be aimed for.

IS THE AIM OF VENTILATION PURELY TO

REPLACE A FAILED OR INACTIVATEDRESPIRATORY PUMP?This may be the goal, for instance, in tetra-plegics, during general anaesthesia with musclerelaxants in patients who have to be sedatedand paralysed to tolerate ventilation or in orderto avoid a raised intracranial pressure due tocoughing. The patient plays a purely passiverole and contributes nothing to the work ofbreathing. Controlled ventilation in which thepatient receives a fixed tidal volume at a pre-determined flow rate and with a fixed frequencyis the mode of choice.

IS THE AIM OF VENTILATION TO SUPPLEMENTTHE PATIENT'S OWN RESPIRATORY ACTIVITY?In this situation the ventilator is more thanpurely an external energy source. It is alsoan additional respiratory control mechanismwhich interacts with the patient's intrinsic res-piratory control system. The respiratory cyclesof both the patient and the ventilator maybe partly or totally synchronised or may beuncoordinated.

Synchronisation of the patient's inspiratoryefforts can trigger ventilator breaths in the assistcontrol and SIMV modes and with pressuresupport. With the latter it can also augment toa variable degree the volume delivered fromthe ventilator, but with volume preset vent-ilators any continuing inspiratory activity willincrease the work of breathing but will notincrease the tidal volume. Triggering and in-spiratory muscle activity after the onset of theventilator's flow can occur separately or to-gether in a single breath.'213An uncoordinated expiratory effort during

the inspiratory phase with a volume presetventilator raises the airway pressure so that airis vented from the system at the pressure limit,thereby reducing the tidal volume. With pres-sure preset machines the predetermined waveform is reached with a lower tidal volume. Thistype of incoordination can be overcome bytriggering the ventilator into expiration by air-

flow rather than inspiratory time, although thistype of flow triggering is only available on somepressure preset ventilators.'4 15

Inspiratory muscle activity during expirationslows the expiratory flow rate (braked ex-piration) which, while it may be beneficial inpreventing airway closure, will trigger anotherinspiration if it leads to a negative airway pres-sure. If this occurs before the previous tidalvolume has been fully exhaled, hyperinflationwill develop (breath stacking), particularly withvolume rather than pressure preset ventilators.There do not appear to be any advantages

from incoordination between the ventilator andthe patient although minor degrees can occurwithout any distress such as in IMV, as long asgas exchange, inspiratory flow rates, respiratoryfrequency, and lung volume remain satis-factory. The converse of incoordination, syn-chrony, is usually - but not always - beneficial.Better alveolar ventilation (as well as respiratorymuscle rest) may be obtained if the patient'srespiration remains passive, stable, and fullysupported by the ventilator.

IS THE AIM OF VENTILATION TO MODIFY THE

PATIENT'S OWN RESPIRATORY ACTIVITY?The patient's respiratory pattern can be modi-fied considerably by the ventilator. Thisimportant but underemphasised concept con-flicts with the usual advice to match the ventil-ator to the patient. The patient can and oftenmust be matched to the ventilator. In its sim-plest form this idea is quite obvious. The res-piratory pattern can be altered through reliefof anxiety and discomfort once adequate ventil-atory support is established. Similarly, res-piration can be modified by normalising thearterial blood gases which reduces the bio-chemical drive to respiration. More subtleinteractions are also possible. The breath tobreath and intrabreath respiratory activity canbe modified by, for instance, altering the in-spiratory flow rate and inspiratory and ex-piratory time, tidal volume and lung volume.The mechanisms probably involve mechano-receptor reflexes from the upper airway andlungs.'6 The microprocessor controls on mod-ern ventilators are becoming increasingly soph-isticated, and although none can approach theflexibility of a spontaneously breathing subject,the scope for influencing the patient's res-piratory activity will undoubtedly increase overthe next few years.

IS THE AIM OF VENTILATION TO TREAT AS WELL

AS TO SUPPORT THE PATIENT?There is no doubt that mechanical ventilationcan support the failing respiratory pump andthat there is a spectrum of interaction betweenthe patient and the machine. What is still un-certain is to what extent, and how, ventilationcan improve the underlying pathophysiologicalchanges which have led to respiratory failure.Most of the effects of ventilation discussedbelow are transient but, with long term noc-turnal ventilation, some must persist longenough to explain the clinical and physiological

758

on February 25, 2021 by guest. P

rotected by copyright.http://thorax.bm

j.com/

Thorax: first published as 10.1136/thx.51.7.756 on 1 July 1996. D

ownloaded from

Techniques in mechanical ventilation: principles and practice

improvements that are seen during the daytime.The most important possibilities are respiratorydrive, respiratory muscle function, respiratorymechanics, and ventilation/perfusion match-ing.

Respiratory driveVentilation can influence respiratory drive by,for instance, altering the arterial Pco, which,by lowering the concentration of bicarbonatein the cerebrospinal fluid, increases the re-sponse to hypercapnia. Relief of sleep dep-rivation also improves the respiratory drivethrough largely unknown mechanisms and, oc-casionally, hypoxic depression ofthe respiratorycentres can also be reversed. The more subtleand rapidly changing interactions between theventilator and the patient's respiratory controlsystem which are mediated through reflex andbehavioural mechanisms have been describedabove.

Respiratory muscle functionThe concept of chronic respiratory muscle fat-igue or incipient fatigue has stimulated effortsto rest the respiratory muscles and reduce thework of breathing.'7 Muscle fatigue has, never-theless, only been demonstrated in a few cir-cumstances and it seems more likely that thecentral respiratory control mechanisms adapttheir output to prevent it from developing. Inany case rest can only be achieved ifnone of theventilator breaths are triggered or augmented,which is unusual if the patient has any residualrespiratory drive and muscle strength. Studiesattempting to rest the respiratory muscles have,not surprisingly, shown no consistent patternof changes. Resting the muscles to relieve oravoid fatigue should be distinguished from theuse ofsedation and muscle paralysis for patientswhose incoordination with the ventilator can-not be controlled in any other way.A further problem with resting the respiratory

muscles is that weaning requires increasingmuscle activity for which it may prove best tohave exercised the muscles in order to conditionor train them rather than risk disuse atrophythrough rest. In practice, the aims of bothrespiratory muscle rest and muscle training areprobably too narrow. The goal should be tooptimise the pattern of respiratory muscle ac-tivity in order to enable the patient initially tobe adequately ventilated by the machine andsubsequently to help to regain as much res-piratory muscle strength and endurance as pos-sible in order to become independent of theventilator. It may be best to rest the respiratorymuscles during the initial stages of an acuteillness or when the patient is systemically un-well, but after this a change in strategy toincrease the work of breathing and improvemuscle performance may assist the weaningprocess.

Respiratory mechanicsMost of the effects of ventilation on respiratorymechanics are transient but a few may persist

after ventilation has been discontinued. A smallimprovement in chest wall and lung compliancemay last for a few hours'8 but other benefitssuch as reversal of abdominal paradoxicalmovement due to diaphragm weakness are lostafter ventilation is discontinued. Alterations inupper airway resistance due to positive pressureduring nasal ventilation are transient but, to-gether with a reduction in dead space, persistwith a tracheostomy as long as this is in usewhether the patient is breathing spontaneouslyor being ventilated.

Ventilationlperf,usion (V/Q) matchingMost ofthe effects ofventilation on V/Q match-ing are also transient. Regional perfusion de-pends on the resistance of the pulmonaryarterial circulation (which is affected by grav-itational forces, lung disease, vascular tone, andlung volume) and cardiac output. An increasein intrathoracic pressure reduces the intra-vascular blood volume, right ventricular inflow,and left ventricular outflow. These problemsare therefore seen if the mean inspiratory pres-sure is high and if the inspiratory time is pro-longed, or if the mean expiratory pressure israised, for instance, with positive end expiratorypressure (PEEP) or continuous positive airwaypressure (CPAP). Intravenous fluids may beneeded to expand the intravascular volumeand oxygen may be added to maintain themyocardial contractility.

It is difficult to manipulate these aspectsof airway, cardiac, and vascular function tooptimise V/Q matching but, in general, this isusually improved by prolonging the inspiratorytime with or without an end inspiratory pause'9to assist equilibration of gases within the alveoliand, ifnecessary, an increase in the tidal volumewith or without PEEP or sighs to prevent airwayclosure.

WHAT ARE THE OTHER ASSOCIATED AIMS OFTREATMENT THAT ARE LINKED TO VENTILATORYSUPPORT?Ventilatory failure is often associated with im-paired bulbar function, especially in neuro-muscular disorders such as motor neuronedisease. The ability to cough, swallow, andspeak may be reduced. In these situations, orif the level of consciousness is impaired, airwayprotection and access to tracheobronchialsecretions with a cuffed tracheostomy or endo-tracheal tube may be as important as ventilatorysupport. In contrast, if the latter is the onlynecessity, a nasal or face mask or a mouth piece(or, if there is upper airway obstruction, anuncuffed tracheostomy tube) may be sufficient.

IS COMPLETE WEANING THE AIM OFVENTILATION?This is usually the end point of ventilation,particularly ifthe patient has undergone surgeryand does not have any chronic respiratory dis-order but long term, usually nocturnal, non-invasive ventilatory support is neverthelessbeing increasingly provided. This is particularly

759

on February 25, 2021 by guest. P

rotected by copyright.http://thorax.bm

j.com/

Thorax: first published as 10.1136/thx.51.7.756 on 1 July 1996. D

ownloaded from

Shneerson

effective in neuromuscular and skeletal dis-orders but is also of some benefit to carefullyselected patients with chronic airflow ob-struction.20 These subjects are, in effect, onlypartially weaned and this long term optionshould be considered during the phase of con-tinuous ventilator dependency if this is pro-longed following an acute illness.

Ventilation in practiceThe optimal ventilator settings for each patientare mainly determined by the metabolic re-quirements, respiratory drive, and mechanics.Normal subjects under general anaesthesia orafter an overdose of a sedative drug, and thosewith little residual respiratory drive or strength- for example, tetraplegics - are easy to ventilatein the controlled mode since their mechanicsare normal. Underinflation can be avoided byincreasing the tidal volume of the ventilator,with or without PEEP, and hyperinflation canbe prevented by reducing the tidal volume andfrequency, avoiding breath stacking, length-ening the expiratory time, and adding PEEPto counteract any autoPEEP. In subjects withsome residual respiratory activity, respiratorymuscle rest may be achieved by reducing therespiratory drive by, for instance, increasing theinspired oxygen concentration or lowering thePco2, increasing the level of pressure supportor tidal volume, and increasing the frequencyof the ventilator. Adjustment of the ventilatorsettings can also promote synchronisation orprovoke incoordination. Synchronisation is, ingeneral, facilitated by a low circuit inspiratoryand expiratory resistance, a low trigger thresh-old with a short delay (which may necessitateflow cycling from expiration to inspiration andin some patients from inspiration to expirationas well), and a sufficiently fast inspiratory flowrate to meet the patient's demands.2'The ventilator can usually be easily adjusted

for patients with neuromuscular and skeletaldisorders despite their reduced chest wall andlung compliance. Assist control or pressuresupport ventilation is preferable to controlledventilation since it allows the patient to triggerinspiration and, in the case ofpressure support,to have some control over tidal volume as well.A high respiratory frequency and low tidalvolume are usually required, but if this is verylow either a reduction in frequency with a largertidal volume or the addition of PEEP canovercome any tendency to basal airway closure.

Similar considerations apply to patients withdiffuse pulmonary disorders such as fibrosingalveolitis, although hypercapnia is usually onlya preterminal complication and ventilatory sup-port is rarely indicated. These patients oftenhave a greatly reduced respiratory system com-pliance and a strong respiratory drive, perhapsdue to stimulation of pulmonary mechano-receptors. The high respiratory drive can be atleast reduced by relieving hypoxia and hyper-capnia and it is important to maximise syn-chronisation between the patient and theventilator. A high mean inspiratory flow rate,low tidal volume, high frequency, and ap-propriate adjustment of the trigger threshold

may enable this to be achieved. If the patientmakes expiratory efforts during inspiration theinspiratory time may need to be shortened,although this may worsen any hypoxia. Al-ternatively, a flow cycled ventilator with pres-sure support may be preferable as long as theventilator can achieve a sufficiently high pres-sure to generate an adequate tidal volume.

Patients with airflow obstruction due to, forinstance, chronic bronchitis or asthma posedifferent problems. Their expiratory flow lim-itation leads to an increase in end expiratoryvolume and pressure (autoPEEP) unless theexpiratory time is sufficient for the previoustidal volume to be completely exhaled. Theexpiratory resistance of the ventilator and inter-face should be minimised by, for example, theuse of a larger tracheostomy tube or a lowresistance expiratory valve. In order to preservean adequate expiratory time the inspiratorytime is shortened and the respiratory frequencyis reduced. Tidal volume is kept small to mini-mise the risk of barotrauma which is related tothe peak inspiratory pressure, although someof this is dissipated along the airways and is notapplied to the alveoli. PEEP may be required toovercome any autoPEEP.22These constraints due to mechanical con-

siderations lead to a narrow range of settingswhich is able to provide adequate alveolar vent-ilation. An important additional problem, how-ever, is the instability of the respiratory pattern.During wakefulness this is due to behaviouralinfluences23 such as anxiety, discomfort, andpain, but during sleep different factors operate.Firstly, there are physiological irregularitiesduring light non-rapid eye movement sleep,usually at sleep onset, and during phasic rapideye movement sleep. Secondly, upper airwayresistance increases and diaphragmatic func-tion is reduced, particularly during rapid eyemovement sleep, so that lung volume changesand chemoreflexes and mechanoreflexes areactivated. Any increase in tidal volume, how-ever, requires a longer expiratory time to pre-vent an increase in end expiratory volume andautoPEEP from developing ifexpiratory airflowlimitation is severe, whereas a reduction in tidalvolume leads to a disproportionate reductionin alveolar ventilation because of the highphysiological dead space in these patients. Theresulting changes in blood gas tensions modifythe patient's respiratory drive and may tend toperpetuate the oscillation of the respiratorypattern outside the narrow range of settingsthat enable adequate ventilation to be achieved.The combination of mechanical limitations,

respiratory instability, and rapidly changingblood gas tensions leads to a constantly fluc-tuating tidal volume, lung volume, respiratoryrate, inspiratory and expiratory times and flowrates, both when the patients are clinicallystable and particularly during either infectiveexacerbations or, in the case of asthma,acute episodes of bronchoconstriction. Syn-chronisation with the ventilator varies frommoment to moment and it may be necessaryto reduce the patient's respiratory drive by,for instance, lowering the metabolic rate (forexample, by treating any infection), relieving

760

on February 25, 2021 by guest. P

rotected by copyright.http://thorax.bm

j.com/

Thorax: first published as 10.1136/thx.51.7.756 on 1 July 1996. D

ownloaded from

Techniques in mechanical ventilation: principles and practice

hypoxia, optimising lung volume, improvingair flow obstruction and considering high fat,low carbohydrate feeding to reduce carbon di-oxide production. It may then be possible tosynchronise the patient's respiratory effort withthe ventilator which will need to be readjustedto take into account the patient's residual res-

piratory drive and mechanics in order to stabil-ise the respiratory rate and inspiratory andexpiratory timing and to sustain an increased

alveolar ventilation. If this cannot be achievedthe alternatives are to accept a high Pco, (con-trolled hypoventilation) or to sedate and para-lyse the patient.

ConclusionMechanical ventilation of patients with res-

piratory failure is rarely a question of applyinga machine to a passive patient. The interactionbetween the ventilator and the patient is com-plex and the properties of the ventilator andits ability to modify the patient's respirationhave to be carefully considered. Clearly definedaims, an understanding of the principles andpractical limitations of the equipment, and a

perceptive evaluation of the patient's changingrequirements are all essential if patients are tobe optimally managed with ventilatory support.

1 Blanch PB, Jones M, Layton AJ, Camner N. Pressurepreset ventilation. Part 1: Physiologic and mechanicalconsiderations. Chest 1993;104:590-9.

2 Bach JR, Sortor SM, Saporito LR. Interfaces for non-invasive intermittent positive pressure ventilatory supportin North America. Eur Respir Rev 1993;3:254-9.

3 Sassoon CSH. Positive pressure ventilation - alternatemodes. Chest 1991;100:1421-9.

4 Manning HL, Molinary EJ, Leiter JC. Effect of inspiratoryflow rate on respiratory sensation and pattern ofbreathing.Am Respir Crit Care Med 1995;151:751-7.

5 Dennison FH, Taft AA, Mishoe SC, Hooker LL, EatherlySB, Beckham RW. Analysis of resistance to gas flow innine adult ventilator circuits. Chest 1989;96: 1374-9.

6 Ferguson GT, Gilmartin M. Carbon dioxide rebreathingduring BiPAPR ventilatory assistance. Am J Respir CritCare Med 1995;151:1126-35.

7 Jounieaux V, Aubert G, Dury M, Delguste P, RodensteinDO. Effects of nasal positive-pressure hyperventilation onthe glottis in normal awake subjects. J Appl Physiol 1995;79:176-85.

8 Jounieaux V, Aubert G, Dury M, Delguste P, RodensteinDO. Effects of nasal positive-pressure hyperventilation onthe glottis in normal sleeping subjects. YApplPhysiol 1995;79:186-93.

9 Smith IE, Shneerson JM. A progressive care programme forprolonged ventilatory failure: analysis of outcome. Br JAnaesth 1995;75:399-404.

10 Tuxen DV. Permissive hypercapnic ventilation. AmJ RespirCrit Care Med 1994;150:870-4.

11 Feihl F, Perret C. Permissive hypercapnia - how permissiveshould we be? Am J Respir Crit Care Med 1994;150:1722-37.

12 Sassoon CSH, Mahutte CK, Te TT, Simmons DH, LightRW. Work of breathing and airway occlusion pressureduring assist-mode mechanical ventilation. Chest 1988;93:571-6.

13 Flick GR, Bellamy PE, Simmons DH. Diaphragmatic con-traction during assisted mechanical ventilation. Chest1989;96:130-5.

14 Branson RD, Campbell RS, Davis K, Johnson DJ. Com-parison of pressure and flow triggering systems duringcontinuous positive airway pressure. Chest 1994;106:540-4.

15 Giuliani R, Mascia L, Recchia F, Caracciolo A, Fiore T,Ranieri VM. Patient-ventilator interaction during syn-chronized intermittent mandatory ventilation - effects offlow triggering. AmJRespir Crit Care Med 1995;151:1-9.

16 Iber C, Simon P, Skatrud JB, Mahowald MW, Dempsey JA.The Breuer-Hering reflex in humans: effects ofpulmonarydenervation and hypocapnia. Am J Respir Crit Care Med1995;152:217-224.

17 Fitting JW. Respiratory muscles during ventilatory support.Eur Respir J 1994;7:2223-5.

18 McCool FD, Mayewski RF, Shayne DS, Gibson CJ, GriggsRC, Hyde RW. Intermittent positive pressure breathingin patients with respiratory muscle weakness: alterationsin total respiratory compliance. Chest 1986;90:546-52.

19 Pillet 0, Choukroun ML, Castaing Y. Effects of inspiratoryflow rate alterations on gas exchange during mechanicalventilation in normal lungs: efficiency of end-inspiratorypause. Chest 1993;103:1161-5.

20 Leger P, Bedicam JM, Comette A, Reybet-Degat 0, Lange-vin B, Polu JM, et al. Nasal intermittent positive pressureventilation: long-term follow-up in patients with severechronic respiratory insufficiency. Chest 1994;105: 100-105.

21 Fabry B, Guttmann J, Eberhard L, Bauer T, Haberthur C,Wolff G. An analysis of desynchronization between thespontaneously breathing patient and ventilator during in-spiratory pressure support. Chest 1995;107:1387-94.

22 Corbridge TC, Hall JB. The assessment and managementof adults with status asthmaticus. Am J Respir Crit CareMed 1995;151:1296-316.

23 Jubran A, Van De Graaff WB, Tobin MJ. Variability ofpatient-ventilator interaction with pressure support vent-ilation in patients with chronic obstructive pulmonarydisease. Am J Respir Crit Care Med 1995;152:129-36.

761

on February 25, 2021 by guest. P

rotected by copyright.http://thorax.bm

j.com/

Thorax: first published as 10.1136/thx.51.7.756 on 1 July 1996. D

ownloaded from