ventilator settings for newborn infants · faster respiratory rates. if the current debate about...

Archives of Disease in Childhood, 1987, 62, 529-538

Controversy

Ventilator settings for newborn infantsC A RAMSDEN AND E 0 R REYNOLDS

Department of Paediatrics, University College London

The selection of appropriate mechanical ventilatorsettings for newborn infants is a controversial topic.The purpose of this review is to summarise theorigins of the controversy and then to concentrateon the interaction between ventilator variables andthe pathophysiology of the lung. As lung mechanicsdiffer widely in different diseases we argue thatcareful consideration must be given to the mechan-ical properties of the lung when deciding on asuitable ventilator regimen for a particular infant.

Origin of early recommendations

Many years ago, one of us made recommendationsfor the ventilation of infants with severe hyalinemembrane disease. These recommendations are stillfollowed by some, but others claim that theypromote pneumothoraces and that better results areobtained if different settings are used-notably,faster respiratory rates. If the current debate abouthow best to set ventilators is to be clarified the originof the early recommendations must be understood.

Mechanical ventilation for newborn infants withrespiratory failure became widespread in the late1960s and early 1970s. Although it was immediatelyclear that the lives of preterm infants with recurrentattacks of apnoea and relatively normal lungs couldbe saved, great difficulty was encountered in infantswith severe hyaline membrane disease (HMD).The conventional way of setting the ventilator at

that time was to select a rate of 60-80 breaths perminute for infants with HMD and if possible tosynchronise the machine with the infant's ownbreathing pattern. At these rates the arterial carbondioxide tension could usually be easily controlledbut the oxygen tension could not, unless very highpeak airway pressures (often 35-50 cm H20) wereused.' The vast majority of the infants died, eitherrapidly of hypoxaemia or more slowly (at about 2weeks of age) from an aggressive form of lungfibrosis, which was termed bronchopulmonarydysplasia.2 3 Evidence was obtained from observations of ventilator variables and autopsy findings

that the main cause of bronchopulmonary dysplasiaat that time was distension and disruption of theterminal airways caused by the use of very high peakairway pressures.3 4Experiments were therefore performed to see if

ways could be found to ventilate babies with severeHMD at lower peak airway pressures.5 6 ModifiedBennett PR2 ventilators (time cycled, pressurelimited, intermittent flow machines) were used, setto provide a plateau of peak airway pressure duringinspiration. The major findings of these studies werethat if a rather slow rate was selected (30 breaths/minute) arterial oxygen tension was higher than at afast rate (an observation also made by Smith et alf)and that the oxygen tension could generally beraised further by manoeuvres that increased meanairway pressure, notably by increasing inspiration:expiration (I:E) ratio. Arterial carbon dioxidetension could most readily be manipulated by smallchanges in ventilator rate or in the differencebetween peak and end expiratory pressures (whichaltered alveolar ventilation).These results were easy to relate to the pathophy-

siology of HMD. As the surfactant deficient lungis both poorly compliant and unstable, it is difficultto inflate and collapses readily during expiration.Atelectasis causing right to left shunting of bloodthrough both intrapulmonary and extrapulmonarychannels is responsible for -most of the deficit ofoxygen uptake.8 Hence improvement of arterialoxygen tension was to be expected when peakairway pressure, I:E ratio, or positive end expiratorypressure were increased-the first strategy openingup collapsed lung units, the second holding themopen for a greater part of each breath, and the thirdretarding alveolar collapse during expiration. Theeffect of slowing the ventilator rate in improvingoxygen tension, though more complex, may in parthave reflected the need for time as well as pressurefor collapsed lung units to reinflate, because viscousand inertial forces must be overcome.4 9 Anystrategy that increases the duration of the inspiratoryphase, such as decreasing rate or increasing I:E ratio

529

on October 9, 2020 by guest. P

rotected by copyright.http://adc.bm

j.com/

Arch D

is Child: first published as 10.1136/adc.62.5.529 on 1 M

ay 1987. Dow

nloaded from

http://adc.bmj.com/

530 Ramsden and Reynolds

(while maintaining other variables constant) should,within limits, promote alveolar inflation, reduceright to left shunt, and improve oxygenation.

These observations on the effects of differentventilator settings led to recommendations for themanagement of infants with severe HMD that weredesigned to avoid the use of very high peak airwaypressures (>25 cm H20). This regimen was char-acterised by slow ventilator rates (30-40/min) andthe use, when required, of prolonged inspiratorytimes (I:E ratio B1:1) as a means for maintaining asatisfactory arterial oxygen tension.'1(1'2 To keeppeak airway pressure and I:E ratio as low aspossible, very high inspired oxygen concentrationswere often used. As soon as recovery started theventilator pressures and inspired oxygen concen-trations were both reduced. The introduction of thisregimen at University College Hospital (UCH) wasassociated with an immediate large increase insurvival that seemed to be attributable to a reductionin deaths from 'aggressive' bronchopulmonarydysplasia.'3Many reservations must be entertained when

applying this regimen today. The experiments onwhich it was based were performed with an intermit-tent flow ventilator on infants who were selected forstudy because they had severe HMD; hence theregimen was recommended only for similar infants.Also the observations on survival predated theregular use of continuous positive airway pressureand positive end expiratory pressure. Perhaps thegreatest change though has been in the populationitself; the thresholds for mechanical ventilation havebecome much less stringent and extremely preterminfants form a far greater proportion of the popula-tion receiving ventilation. Nevertheless, it is in-teresting that at UCH, where we have used thisregimen routinely for severe HMD since 1970, thechange in population does not seem to have beenassociated with any appreciable change in theincidence of pneumothorax the complication cur-rently under most debate. Between 1979 and 1983the incidence was 21% in 88 inborn infants withgestation of 24 to 32 weeks who were ventilated forHMD. (Ramsden CA, Stewart AL. Unpublisheddata.)The controversy

Some authors have experienced high incidences ofpneumothorax-up to 50%-in infants ventilatedwith slow rates and long inspiratory times,14 i5 andseveral publications have implied that the use offaster rates and shorter inspiratory times is associ-ated with better results, including fewer cases ofpneumothorax. 14-22Three studies have been particularly influential in

popularising this view, but each suffers from prob-lems of experimental design. 118 Bland et al re-ported only two cases of pneumothorax in 24 infantsventilated for HMD at 60-110/min but had nocontrol group. 6 Spahr et al found a trend towardsfewer air leaks in infants with HMD ventilated withI:E ratios of 1:2 than in infants ventilated with ratiosof 2:1, but no account was taken of the severity ofthe disease-infants were ventilated for an averageof 80 hours with an I:E ratio of 2:1, even though aninspired oxygen concentration of less than 50% wasrequired for 60% of that time.17 Heicher and hercolleagues reported that in a consecutive series of102 infants requiring ventilation for a variety ofrespiratory disorders pneumothoraces developed inonly 14% of those ventilated at 60/min, significantlyless than the 35% incidence in infants ventilated at20-40/min. i The maximum peak airway pressureallowed in infants ventilated at the fast rate,however, was 30 cm H2O compared with 40 cm H2Ofor the slower rate. More recently, Pohlandt et alhave reported the preliminary results of a largemulticentre trial comparing slow rates and longinspiratory times with fast rates and short inspiratorytimes.'9 A lower incidence of pneumothorax wasfound when fast rates were used, but no account wastaken of the diagnoses.These studies illustrate very well two fundamental

problems of trial design that lie at the heart of thepresent debate about ventilator settings. The first islack of appreciation that the regimen summarisedearlier was developed specifically for infants withsevere HMD. The use of slow rates and longinspiratory times in most other respiratory disordersor in mild or recovering HMD where the lung isfairly compliant and stable is certain to cause severehyperinflation with disastrous results-notably, lungrupture and obstruction of the circulation.'2 Thesecond and closely related problem is the enrolmentof study groups unselected either by respiratorydiagnosis or by the severity of the illness.We believe that before further trials are under-

taken it is vital that consideration is given in theirdesign to the relation between ventilator variablesand lung pathophysiology, the most crucial aspect ofwhich is the relation between the expiratory time ofthe ventilator and the expiratory time constant ofthe respiratory system.

Time constant of the respiratory system

During conventional mechanical ventilation lungdeflation usually occurs passively. As the expiratoryvalve of the ventilator opens gas flows from theinfant's lungs into the ventilator circuit propelled bythe pressure gradient between the alveolar lumen



j.com/

Arch D


ay 1987. Dow

nloaded from

http://adc.bmj.com/

and airway, which is generated by the elastic recoilof the lungs and chest wall.The time taken for lung deflation depends on the

magnitude of the elastic recoil (the inverse ofcompliance) and on the resistance of the airways.Decreased compliance shortens the time taken andincreased airway resistance lengthens it.The time required for expiration is therefore

directly related to compliance and resistance andcan be expressed by the following equation:

Vt _ tV-=e RC

where Vt is the volume remaining in the lung at timet after the onset of expiration, Vo is the total volumeexhaled when expiration is complete, and e is a

constant with the value 2-7183. R is airway resist-ance and C is compliance of the lungs and chest wall(see Appendix for derivation of this relation).The product of compliance and resistance is often

referred to as the time constant of the respiratorysystem (TRS) and is measured in units of time(seconds). TRS provides an index of the timenecessary for deflation to occur, and its relevance toventilation becomes apparent if we solve the aboveequation for various durations of expiration. Forexample, when the duration of expiration 't' equalsTRS then:

Vt -t VtV=e tRS or -=2-7183-'=0-37.

Thus after the passage of a single time constant 37%of the tidal volume remains to be expired. Similarly,we can calculate that after two, three, four, or fivetime constants have passed the proportion remain-ing is 13-5, 5-0, 1-8, and 0-7%, respectively.The value of TRS sets a limit to the minimum

expiratory time that can be used without causing gastrapping. For practical purposes, a time equivalentto at least three time constants must be allowed ifexpiration is to be reasonably (95%) complete.

Time constant and disease state

As TRS is determined by the compliance and resist-ance of the respiratory system it varies widelyaccording to the nature and severity of the respira-tory illness.

If we use published values of R and C2>26 we can

estimate the expected value of TRS for variousrespiratory disorders. It may be as short as 0-05seconds in HMD (compliance 1 ml/cm H20,resistance 50 cm H20/llsec), as long as 025seconds in infants with normal lungs (compliance -

5 mllcm H20, resistance 50 cm H20/l/sec), andconsiderably longer in those with airways obstruc-tion. For example, Grunstein et al recently esti-

Ventilator settings for newborn infants 531

mated mean TRS to be 0-5 seconds in infants withchronic lung disease.26 The time that must beallowed for expiration to occur during mechanicalventilation is therefore crucially dependent on thedisease from which the infant suffers, varying morethan fivefold between different diseases.

Positive end expiratory pressure (PEEP) and'inadvertent PEEP'

If a ventilator expiratory time of less than 3XTRS isused-for example, <075 sec for normal lungs-substantial gas trapping will inevitably occur duringexpiration. The effect of this is quite similar to theeffect of applying PEEP and is often referred to as'inadvertent PEEP'. Either measure will hold lungvolume above the relaxed functional residualcapacity-the functional residual capacity at zeroinflation pressure-and may improve oxygenation ifthe relaxed capacity is abnormally low.

Despite this potentially beneficial effect (whichnewborn infants apparently set out to achieve forthemselves during spontaneous breathing. See be-low.), we believe that inadvertent PEEP representsthe major danger of employing fast rates duringmechanical ventilation. Inadvertent PEEP is addi-tive to the applied PEEP and carries no theoreticaladvantage over increasing the level of appliedPEEP. Unlike applied PEEP, its magnitude canonly be measured by sophisticated techniques, and itmay be quite considerable: Simbruner and hiscolleagues have recently shown levels of inadvertentPEEP as high as 4-7 cm H20 in infants ventilatedfor a variety of respiratory disorders at 30-40/minand with expiratory times of not less than 0O8seconds.27 Furthermore, because the value of TRS iSnot static throughout the course of a respiratoryillness, the level of inadvertent PEEP will varyconsiderably, while ventilator settings remain un-changed. During the recovery phase of HMD, forexample, the use of a fast rate at a time when the TRSis increasing rapidly may cause a dramatic rise ininadvertent PEEP with the risk of carbon dioxideretention, pneumothorax, and compression of thepulmonary circulation.12

Expiratory time

Although it is tempting to use published values ofTRS to estimate the minimum expiratory time thatcan be used in various respiratory disorders withoutcausing air trapping, considerable caution has to beexercised as several factors may alter the effectivevalue of TRS. For example, TRS iS increased by theadditional resistance of the endotracheal tube,which is both variable (dependent on tube diameter



j.com/

Arch D


ay 1987. Dow

nloaded from

http://adc.bmj.com/


and design) and unpredictable (due to the accumula-tion of secretions).28 29 The presence of PEEP(inadvertent or applied), on the other hand, mayshorten the time constant by decreasing resistance(splinting airways open) and decreasing compliance(by moving the tidal flow on to a higher and flatterpart of the pressure-volume curve). The mechanicalbehaviour of the ventilator must also be considered.In particular, the exhalation valve takes a finite timeto open, during which resistance is high and expira-tion is slowed: the opening time is only 0-04-0-07seconds with the Sechrist or Bear Cub BP 2001machines but can be as much as 0-28 seconds withthe Baby Bird.3"

Despite these reservations, it is sobering to makesome rough calculations. The value of TRS in anintubated infant with normal lungs is about 0-25seconds. Ventilated at 60 breaths/min with an L:Eratio of 1:1, almost 15% of the tidal volume will betrapped in expiration; at 100/min the proportionincreases to 30%. The presence of secretions in theendotracheal tube can double respiratory systemresistance;28 in our last example this would increasethe trapped proportion to 55%. Under such cir-cumstances inadvertent PEEP would be very great.The crucial effect of lung disease on expiratorytime is evident when similar calculations are per-formed for infants with respiratory illnesses, usingvalues of TRS given earlier. An immediate visualmeaning to TRS (as well as useful diagnostic infor-mation) is given by watching the rate of expiration inan ill infant: the lungs of an infant with severe HMDseem to collapse very rapidly, whereas those of aninfant with meconium aspiration may seem to bealmost fixed in inspiration.

Inspiratory time

Similar principles can be applied when consideringthe effect of manipulating ventilator inspiratorytime. The factors affecting lung inflation are,however, more complex than for deflation and suchan approach is rather crude. Nevertheless, twoissues are quite important.The first is that airway resistance is lower during

inspiration than during expiration. Hence the in-spiratory time necessary for lung inflation to becomplete is considerably less than the correspondingexpiratory time. During spontaneous breathing thiseffect is probably quite small, expiratory resistancebeing only 15-20% greater than the inspiratoryresistance. Perez Fontan and his colleagues haverecently reported, however, that during mechanicalventilation the combined mean expiratory resistance(respiratory system plus endotracheal tube) may beas much as 4-5 times greater than the mean

inspiratory resistance.3" This effect may be ex-plained in part by airways being splinted open by theapplied inspiratory pressure. Such a striking differ-ence between inspiratory and expiratory resistancefavours gas trapping when the ventilator rate isincreased; lung inflation will be complete withinspiratory times that are considerably less than theminimum effective expiratory time.The second issue is that where atelectasis is

present, as in HMD, inspiratory times considerablylonger than those predicted from TRS may berequired to overcome the forces that resist re-expansion of collapsed lung units (as discussedabove).

Inhomogeneity of the lung

So far we have assumed that the lung behaves as if itwas completely homogeneous-each lung unit oper-ating in parallel, with an identical TRS. For manypractical purposes this assumption seems warrantedas a linear relation between volume and flow,implying a single value for TRS (see Appendix), hasbeen found in normal infants and infants withHMD.24 25 As ventilatory rate increases, however,progressive inhomogeneity of TRS becomes appar-ent, even in normal lungs, as shown by Helliesen eta13' and Olinsky et al.32 There are two importantconsequences. Firstly, as inspiratory time shortenslung units with relatively long values for TRS becomeless ventilated, leading to ventilation-perfusion im-balance, and secondly, as expiratory time shortensgas trapping develops in units with long values forTRS.

Gross inhomogeneity of the lung is present incertain disease states-notably, meconium aspira-tion, which causes widespread uneven airwaysobstruction-and in chronic lung disease, whereboth resistance and compliance vary widely through-out the lung, some areas being fibrotic and othersemphysematous. Clearly, no single value of TRS canbe assumed when considering how best to ventilatethese infants.

For infants with severe meconium aspiration, ithas been suggested that advantage can be taken ofthe inhomogeneity of TRS. 12 If a very short inspiratorytime is used lung units with fairly short values forTRs-that is, the relatively less obstructed areas oflung-will be preferentially ventilated, thus avoid-ing, partly at least, the dangerous complication ofsevere gas trapping distal to the obstructions.

Spontaneous breathing

Infants often continue to make active respiratoryefforts during mechanical ventilation, and several



j.com/

Arch D


ay 1987. Dow

nloaded from

http://adc.bmj.com/

studies suggest that this activity increases the risk ofpneumothorax33 and cerebral haemorrhage.34 Aparticular pattern of interaction between infant andventilator, so called 'active expiration', has re-cently been implicated as responsible,33 though themechanism by which this interaction causespneumothorax remains uncertain.35 In a populationof infants with this behaviour Greenough et al foundthat muscle relaxation with pancuronium almostcompletely prevented pneumothorax (one case in 11infants), whereas all 11 comparable controls de-veloped this complication.33

It is often suggested that faster ventilator ratespromote 'synchrony' between infant and ventilator.While this may be a common clinical impression,surprisingly little experimental evidence is availableto confirm it. Both Greenough et al and Field et alhave shown some increase in the time spent inapnoea or synchronous breathing when fast ventila-tor rates were employed, but neither group attemp-ted to keep the major variable affecting respiratorydrive constant-namely, the carbon dioxidetension.21'22 Intuitively, it may seem reasonable toattempt to match ventilator timing to the infant'sown spontaneous ventilatory cycle. Given that thecontrol of breathing, however, depends on complexinteractions between reflex responses-for example,Hering-Breuer reflex, carbon dioxide tension, sleepstate, and behavioural factors-there may be atemptation to expect no more than transitorysuccess.There is increasing evidence that the spon-

taneously breathing infant (like the mouse37) uses arapid respiratory rate as a means of preventingairway closure in expiration.36 By employing a rapidrate and a very short expiratory time (considerablyless than 3X TRS), complete expiration is prevented,a larger functional residual capacity is maintained,and oxygenation is improved. Attempts to achievethe same effect with a mechanical ventilator, asdiscussed above under 'inadvertent PEEP', arefraught with danger. Both the applied force (in-advertent PEEP) and its desired effect (raisedfunctional residual capacity) are difficult to measure.The situation is quite unlike that of the spontaneouslybreathing infant whose lungs are liberally endowedwith stretch receptors providing the feedback infor-mation necessary to modulate respiratory timing inaccordance with the changing lung mechanics.

Conclusions

The arguments and calculations presented here areno more than illustrative, but we suggest that theyshow the difficulties of selecting appropriate ventila-tor settings for newborn infants and the folly of

Ventilator settings for newborn infants 533

performing clinical trials that compare ventilatorregimens in mixed populations of infants. Theresults of these trials must depend as much on thecharacteristics of the population as on the treatmentapplied: important beneficial effects in infants withone type of illness may be completely submerged byadverse effects in another. We contend that only farmore carefully designed and disease specific trialswill provide any useful answers to questions aboutwhat type of ventilation to use or how best to set theventilator variables. Wherever possible, measure-ments of lung mechanics, including TRS, as well asblood gases, should be employed in these trials sothat the true effects of ventilation can be assessed.Our own practice, pending further information, is

based on the principles outlined above. We decideon what pattern of ventilation to use according tothe respiratory problem from which the infantsuffers; in so doing we hope to obtain satisfactoryblood gases by tailoring the ventilator settings to suitthe infant's lung mechanics. In general, we use thesame disease specific guidelines as described indetail in 1979.12 We rarely use rates greater than40/min. We watch the infant's chest movementscarefully to be sure that the lung is inflating and toassess whether expiration seems complete beforethe next ventilator breath: a visual clue about theduration of TRS can thus be obtained and the risks ofinadvertent PEEP and gas trapping minimised. Weoften use pancuronium for muscle relaxation in thelarger infants but are wary of it in the smallestones.38One important issue not touched on here is the

potential place of ventilation of newborn infants byhigh frequency oscillation:39 it is at present too earlyto draw firm conclusions about this very promisingtechnique.

Appendix

According to the equation of motion of the respira-tory system

VP=-+RVC (1)

where P is the distending pressure (airway pressureduring intermittent positive pressure ventilation), Cthe compliance, and R the resistance of the lungsand chest wall. V represents the volume abovefunctional residual capacity and V the rate of gasflow measured at the airway opening.At the onset of a passive expiration P falls to zero

and equation (1) can be rearranged to giveV-= _RCV

orV=-VTRS (2)



j.com/

Arch D


ay 1987. Dow

nloaded from

http://adc.bmj.com/


This relation is exploited in single breath studies oflung mechanics. If flow is plotted against expiredvolume a straight line is obtained, from which TRScan be derived. Note, however, that a linear relationbetween V and V will only be evident where thelung is homogeneous with a single value for TRS.

Integration of equation (2) yieldsVt _ tV =e RC

from which the volume of gas still present in thelungs, Vt, at any time, t, after the onset ofexpiration can be calculated (Vo is the total volumeexpired at the end of a complete expiration).References

Adamson TM, Collins LM. Dehan M, Hawker JM, ReynoldsEOR, Strang LB. Mechanical ventilation in newborn infantswith respiratory failure. Lanicet 1968;ii:227-31.

2 Northway WH Jr, Rosan RC, Porter DY. Pulmonary diseasefollowing respiratory therapy of hyaline membrane disease:bronchopulmonary dysplasia. N Engl J Med 1967;276:257-68.Hawker JM, Reynolds EOR, Taghizadeh A. Pulmonary surfacetension and pathological changes in infants dying after respiratortreatment for severe hyaline membrane disease. Lancet 1967;ii:75-7.

4 Taghizadeh A, Reynolds EOR. Pathogenesis of bronchopul-monary dysplasia following hyaline membrane discase. Am JPathol 1976;82:241-58.

5 Reynolds EOR. Effect of alterations in mechanical ventilatorsettings on pulmonary gas exchange in hyaline membranedisease. Arch Dis Child 1971;46:152-9.Herman S, Reynolds EOR. Methods for improving oxygenationin infants mechanically ventilatcd for scvcre hyaline membranedisease. Arch Dis Child 1973;48:612-7.

7 Smith PC, Schach MS, Daily WJR. Mechanical ventilation ofnewborn infants: II. Effects of independent variation of rate andprcssure on arterial oxygenation of infants with respiratorydistress syndrome. Anesthesiology 1972:37:498-502.

8 Strang LB. Neoniatal respiration. Oxford: Blackwell, 1977.9 Agostoni E, Taglietti A, Agostoni F, Setnikar I. Mechanical

aspects of the first brcath. J Appl Phvsiol 1958;13:344-8."' Reynolds EOR. Pressure waveform and ventilator settings for

mcchanical vcntilation in severe hyaline membrane disease. IntAnesthesiol Clin 1974;12:259-80.

" Reynolds EOR. Management of hyaline membrane discase. BrMed Bull 1975;31:81-4.

12 Reynolds 0. Ventilator therapy. In: Thibeault DW, Greg-ory GA, eds. Neonatal pulmonary care. Menlo Park: Addison-Wesley, 1979:217-36.

3 Reynolds EOR, Taghizadch A. Improved prognosis of infantsmechanically ventilated for hyaline membrane disease. Arch DisChild 1974;49:505-15.

4 Primhak RA. Factors associated with pulmonary air leak inpremature infants receiving mechanical ventilation. J Pediatr1983;102:764-8.

I5 Tarnow-Mordi WO, Narang A, Wilkinson AR. Lack of associa-tion between barotrauma and air leak in hyaline membranedisease. Arch Dis Child 1985;60:555-9.

16 Bland RD, Kim MH, Light MJ, Woodson JL. High frequencymechanical ventilation in severe hyaline membrane disease. CritCare Med 1980;5:275-80.

17 Spahr RC, Klein AM, Brown DR, MacDonald HM, Holz-man IR. Hyaline membrane disease: a controlled study ofinspiratory to expiratory ratio in its management by ventilator.Am J Dis Child 1980;134:373-6.

18 Heicher DA, Kasting DS, Harrod JR. Prospective clinicalcomparison of two methods for mechanical ventilation of

neonates: rapid rate and short inspiratory times versus slow rateand long inspiratory time. J Pediatr 1981;98:957-61.

'9 Pohlandt F, Bernsau V, Feilen K-D, et al. Reduction ofbarotrauma in ventilated neonates by increase in ventilationfrequency-first results of a prospective collaborative andrandomized trial of two different ventilatory techniques. PediatrRes 1986;19:1()77.

2(1 Field D, Milner AD, Hopkin IE. High and conventional rates ofpositive pressure ventilation. Arch Dis Child 1984;59:1151-4.

21 Field D, Milner AD, Hopkin IE. Manipulation of ventilatorsettings to prevent active expiration against positive pressureinflation. Arch Dis Child 1985;60:1036-40.

22 Greenough A, Morley CJ, Pool J. Are fast rates an effectivealternative to paralysis? Pediatr Res 1986;19:1077.

23 Mortola JP, Fisher JT, Smith B, Fox G, Weeks S. Dynamics ofbreathing in infants. J Appl Physiol 1982;52:1209-15.

24 Thomson A, Silverman M. Single-breath measurement of lungmechanics in very low birthweight infants. Crit Care Med1985 :13:4-8.

25 Le Souef PN, England SJ, Bryan CA. Passive respiratorymechanics in newborns and children. Am Rev Respir Dis1984;128:552-6.

26 Grunstein MM, Inscore SC. A new method to assess respiratorymechanics in infants with pulmonary disease. Pediatr Res 1986;20:472.

27 Simbruner G. Inadvertent positive end-expiratory pressure inmechanically ventilated newborn infants: detection and effecton lung mechanics and gas exchange. J Pediatr 1986;108:589-95.

2X Prendiville A, Thomson A, Silverman M. Effect of tracheo-bronchial suction on respiratory resistance in intubated pretermbabies. Arch Dis Child 1986;61:1178-83.

29 LeSouef PN, England SJ, Bryan AC. Total resistance of therespiratory system in preterm infants with and without anendotracheal tube. J Pediatr 1984;104:108-11.

30 Perez Fontan JJ, Heldt GP, Targett RC, Willis MM,Gregory GA. Dynamics of expiration and gas trapping in rabbitsduring mechanical ventilation at rapid rates. Crit Care Med1986;:14:39-47.

3' Helliesen PJ, Cook CD, Friedlander L, Agathon S. Studies ofrespiratory physiology in children: 1. Mechanics of respirationand lung volumes in 85 normal children 5 to 17 years of age.Pediatrics 1958;22:80-93.

32 Olinsky A, Bryan AC, Bryan MH. A simple method ofmeasuring total respiratory system compliance in newborninfants. S Afr Med J 1976;50:128-30.

33 Greenough A, Wood S, Morley CJ, Davis JA. Pancuroniumprevents pneumothoraces in ventilated premature babies whoactively expire against positive pressure inflation. Lancet 1984i:1-3.

34 Perlman JM, Goodman S, Kreusser KL, Volpe JJ. Reduction inintraventricular hemorrhage by elimination of fluctuating cer-ebral blood-flow velocity in preterm infants with respiratorydistress syndrome. N Engl J Med 1985;312:1353-7.

35 Ramsden CA. Active expiration or synchrony? Arch Dis Child1986;61:820.

36 Kosch PC, Stark AR. Dynamic maintenance of end-expiratorylung volume in full term infants. J Appl Physiol 1984;57:1126-33.

37 Vinegar A, Sinnett EE, Leith DE. Dynamic mechanismsdetermine functional residual capacity in mice, Mus musculus.J Appl Physiol 1979;46:867-71.

38 Reynolds EOR, Hope PL, Whitehead MD. Muscle relaxationand periventricular hemorrhage. N Engl J Med 1985;313:955-6.

39 Marchak BE, Thompson WK, Duffty P, et al. Treatment ofRDS by high-frequency oscillatory ventilation: a preliminaryreport. J Pediatr 1981;99:287-92.

Correspondence to Dr C A Ramsden Department of Paediatrics,University College London, The Rayne Institute, UniversityStreet, London WC1E 6JJ.



j.com/

Arch D


ay 1987. Dow

nloaded from

http://adc.bmj.com/

ventilator settings for newborn infants · faster respiratory rates. if the current debate about...

Documents