Early Human Developmemt. 25 ( 1991) 69-13
69
Ekevier Scientific Publishers Ireland Ltd.
EHD 01134
Spontaneous respiratory effort during mechanical ventilation in infants with and without acute respiratory distress M.F. Hird and A. Greenough Dept. qj’Child Health. King’s College Hospital. London (U.K.)
(Received 2 May 1990; revisions received 3 September 1990 and 25 January 1991; accepted 29 January 1991)
Summary
Respiratory interactions of 27 ventilated preterm infants were recorded daily during the first 14 days of life to assess the effect on respiratory efforts of recovery from acute respiratory distress syndrome (RDS). Active expiration and persistent asynchrony only occurred during acute RDS (P c 0.01). Throughout the 1Cday period, in the majority of infants making respiratory efforts, a ventilator rate could be found from a standard sequence 30,60,90, 120 breathsjmin which provoked a synchronous interaction, but with increasing postnatal age apnoea became more common (P < 0.01). We conclude that the preterm infants’ spontaneous respiratory efforts are a less important influence on the outcome of mechanical ventilation following recovery from acute RDS. respiratory interaction; mechanical ventilation; respiratory distress syndrome.
Introduction
Preterm neonates frequently make respiratory efforts during mechanical ventilation which can be divided into distinct interactions [5]. The type of respiratory interaction provoked and hence the outcome of mechanical ventilation [5] appears to be dependent on the infant’s spontaneous respiratory timing. Preterm neonates breathe Correspondence CO:Dr A. Greenough, Dept. of Child Health, King’s College Hospital, London SE5 9RS.
U.K. 0378-3782/91/$03.50 0 1991 Elsevier Scientific Publishers Ireland Ltd. Published and Printed in Ireland
70
at rapid respiratory rates, with short inspiratory and expiratory times [4] and high frequency positive pressure ventilation (HFPPV, rates equal to or in excess of 60 breaths/min) closely mimics this inducing synchrony which improves oxygenation [8]. Lengthening inflation time, however, provokes active expiration [ 1,2,6], increasing the risk of pneumothorax [lo]. As a consequence, during acute respiratory distress syndrome (RDS) HFPPV improves outcome compared to ventilation employing rates less than or equal to 40 breathsimin [3,11]. The studies [ 1,2,4-6,8] which have examined the infant’s spontaneous interaction with mechanical ventilation and their respiratory timing have all been carried out in the first week of life and amongst infants suffering from acute respiratory distress. Increasing survival of VLBW infants [9] has meant that many infants require mechanical ventilation even after recovery from RDS and they remain ventilated at least into the second week of life. It cannot be concluded that HFPPV would necessarily be the optimum form of ventilation in these older infants. As RDS resolves, lung compliance may improve which would lengthen the time constant of the respiratory system. Such a change in lung function could result in lengthened inspiratory and expiratory times and thus influence the spontaneous respiratory interactions provoked at different ventilator rates. The aim of this study was to determine if recovery from RDS did indeed influence the nature of the infant’s respiratory efforts during positive mechanical ventilation. Patients and Methods
Consecutive VLBW infants were entered into the study if they required mechanical ventilation at rates > 30 breathsjmin for at least 14 days from birth. Daily measurements were made in two study periods, days U representing acute RDS and from days 5 to 14 representing non-acute respiratory distress. Respiratory measurements were made using the system previously described in detail [4]. Briefly, a pneumotachograph was inserted between the endotracheal tube and ventilator circuit. This recorded flow changes, both due to the infant and the ventilator. This signal was electronically integrated to give volume. Ventilator pressure changes were recorded proximal to the pneumotachograph. Oesophageal pressure changes were recorded using an oesophageal balloon. Flow, volume, oesophageal and ventilator pressure changes were recorded simultaneously on a Gould polygraph. Measurements were made for 20-min periods and only when the infant was stable with blood gases within the clinically acceptable range (pH 7.25-7.4; Pco2 35-45 mmHg; PO* 45-70 mmHg). The infant’s respiratory interaction was determined as that which occurred on more than 80% of positive pressure inflations. The infant’s respiratory interactions were classified as previously defined [7]: (1) apnoea; (2) synchrony; (3) asynchrony. At the initial study the infant’s respiratory interaction was studied at a series of ventilator rates (30, 60, 90 and 120/min). During subsequent study periods infants were studied only at the ventilator rate they were currently receiving but re-studied at a series of rates if they were found to be asynchronous or paralysis had recently been stopped. In the study periods when infants were ventilated at a series of rates
71
they were described as synchronous if synchrony was provoked by at least one of the ventilator rates and subsequently ventilated at the slowest rate at which the synchronous interaction was achieved [6]. Infants were described as asynchronous if they remained asynchronous at all the ventilator rate sequences, they were then paralysed and paralysis continued until the infant’s respiratory distress had improved so that peak pressures could be reduced to 18 cmH20 and rates reduced below 40 breaths/min [lo]. Apnoeic infants were ventilated at the slowest ventilator rate in the sequence [6]. Twenty-seven infants were enrolled into the study with a median gestational age of 27 weeks (range 23-31) and birthweight of 778 g (range 5061500). All of the infants had been ventilated from birth for RDS. Respiratory support was provided by Sechrist ventilators via oral endotracheal tubes. Ethical permission for this study was granted by the King’s College Hospital Ethics Committee. Statistical
analysis
Differences in the respiratory interaction provoked at different postnatal ages were assessed for statistical significance using Fisher’s exact test. Results
During the first four days of life the majority of infants were making respiratory efforts, only 4 of 27 being apnoeic on day 1 (Table I). With increasing postnatal age the proportion of infants without respiratory efforts increased, being greater than 50% from day 7 onwards (day l-4 compared to day 1l-14, P < 0.01). Persistent asynchrony was only seen in the first two days of life, the difference between the
TABLE
I
Spontaneous
Day
respiratory
interactions
Apnoea
in ventilated
infants
(n = 27)
Synchrony
Asynchrony
I 2
4 4
15 II
8” 4”
3 4 5 6 7 8 9 10
4 4 9 11 14 14 14 14
II I1 8 9 IO 10 IO II
II 12 13 14
15 15 15 15
II I1 II II
0 0 0 0 0 0 0 0 0 0 0 0
Paralysed 8 8+4 12 I2 10 7 3 3 3 2 I I I
1
“On day I, 8 infants were asynchronous despite rate manipulation and on day 2, 4 additional became persistently asynchronous, these infants were paralysed on day 1 or 2, respectively.
infants
72
occurrence of apnoea at that time and the rest of the period was highly significant (P < 0.001). All 12 infants persistently asynchronous, despite being studied at the rate sequence, were paralysed. Once pancuronium was discontinued, none of these 12 infants were asynchronous despite being studied at the whole rate sequence, 9 of the infants were apnoeic and 3 had synchronous respiratory efforts. Throughout the 14-day study period, in the majority of infants making respiratory efforts, synchrony could be achieved by increasing the ventilator rate to 60 breaths/min or beyond (Table I). Discussion
In this study we have demonstrated that the majority of infants, even including those born very preterm, were making respiratory efforts during the acute state of their illness. We had previously found [5] that small, very sick premature infants rarely make spontaneous respiratory efforts. In that study [5], however, infants were only studied at one ventilator rate and in the present trial infants were studied on day one at a series of rates; this [8] seems likely to explain the difference in the results of the two studies. Our results have confirmed [3,5] that respiratory interaction usually remains stable at a constant ventilator rate. Except in the first two days of life, infants who were ventilated at a rate which provoked synchrony, remained synchronous (the majority) or became apnoeic, until ventilator rate was altered to be reduced below 30 breaths/min. Our results demonstrate that with increasing postnatal age there is a change in the nature of the preterm infant’s respiratory efforts, asynchrony disappearing and apnoea becoming more common (Table I). These changes may be due to recovery from acute RDS. Infants who are asynchronous have previously been demonstrated to have the most severe RDS [2]. Yet, even in this group, their respiratory status had recovered sufficiently during the second study period (days 5-14) for their ventilator rates and pressure to have been reduced sufficiently so that treatment with pancuronium could be discontinued, indicating recovery from acute RDS. An alternative explanation for the increasing proportion of infants who were apnoeic at the higher postnatal ages is an effect of pancuronium. Although certain of the apnoeic infants had breathed synchronously with the ventilator prior to day 5, the majority had been paralysed during the acute stage of their illness. Once recovery from RDS had commenced, asynchrony could not be provoked despite studying infants at the whole rate sequence. Asynchrony is provoked at low ventilator rates amongst infants with the stiffest lungs [2]. The failure to provoke asynchrony once recovery from RDS had commenced is thus likely to be due to the improvement in lung function occurring at this time. The lack of asynchrony, even at long inspiratory times, after day 4 of life, may explain the low incidence of pneumothorax during recovery from acute RDS [ 141. This study demonstrates that in the second week of life, it is not necessary to manipulate ventilator rate to avoid active expiration [ 1,4,6]. We must therefore conclude that spontaneous respiration seems less likely to cause an airleak during
73
mechanical ventilation in the second rather than the first week of life [5,10]. Although synchrony could be provoked at a standard fast ventilator rate in infants who made respiratory efforts in the second week of life, this interaction has not been proved to improve oxygenation following recovery from RDS. Increasing numbers of infants ventilated beyond the first week of life now make it imperative to determine, by blood gas measurements at a series of ventilator rates, optimum settings for such infants. References 1 2 3 4 5 6 I 8
9 10
I1 12
13 14
Field, D., Milner, A.D. and Hopkin, I.E. (1984): High and conventional rates of positive pressure ventilation. Arch. Dis. Child., 59, 1151-1154. Greenough, A., (1988): The premature infant’s respiratory response to mechanical ventilation. Early Hum. Dev., 17, l-5. Greenough, A., Dixon, A. and Roberton, N.R.C. (1984): Pulmonary interstitial emphysema. Arch. Dis. Child., 59, 1046-l 105. Greenough, A., Greenall, F. and Gamsu, H. (1987): Synchronous respiration - which ventilator rate is best? Acta. Paediatr. &and., 76, 713-718. Greenough, A., Morley, C.J. and Davis, J.A. (1983): Interaction of spontaneous respiration and artificial ventilation. J. Pediatr., 103, 769-773. Greenough, A., Morley, C.J. and Pool, J. (1986): Fighting the ventilator - are fast rates an effective alternative to paralysis? Early Hum. Dev., 13, 189-194. Greenough, A. and Greenall, F. (1988): Observation of spontaneous respiratory interaction with artificial ventilation. Arch. Dis. Child., 63, 168-171. Greenough, A., Pool, J., Greenall, F., Morley, C.J. and Gamsu, H. (1987): Comparison of different rates of artificial ventilation in preterm neonates with the respiratory distress syndrome. Acta. Paediatr. &and., 76, 70&12. Greenough, A. and Roberton, N.R.C. (1985): Morbidity and mortality in neonates ventilated for the respiratory distress syndrome. Br. Med. J., 290, 597-600. Greenough, A., Wood, S., Morley, C.J. and Davis, J.A. (1984): Pancuronium prevents pneumothoraces in ventilated premature infants who actively expire against positive pressure ventilation. Lancet, i, l-3. Morley, C.J. and Greenough, A. (1990): Respiratory compliance measurements in very premature babies treated with artificial surfactant (ALEC). Arch. Dis. Child., in press. Heicher, D.A., Kastings, D.S. and Richards, J.R. (1981): Prospective clinical comparison of two methods for mechanical ventilation of neonates: rapid rate and short expiratory time. J. Pediatr., 98, 957-961. Hird, M., Greenough, A. and Gamsu, H. (1990) Gas trapping during high frequency pressure ventilation using conventional ventilators. Early Hum. Dev., 22, 51-56. Tarnow-Mordi, W.O., Narang, A. and Wilkinson, A.R. (1985): Lack of association barotrauma and airleak in hyaline membrane disease. Arch. Dis. Child., 60, 555-559.
positive between
69
Ekevier Scientific Publishers Ireland Ltd.
EHD 01134
Spontaneous respiratory effort during mechanical ventilation in infants with and without acute respiratory distress M.F. Hird and A. Greenough Dept. qj’Child Health. King’s College Hospital. London (U.K.)
(Received 2 May 1990; revisions received 3 September 1990 and 25 January 1991; accepted 29 January 1991)
Summary
Respiratory interactions of 27 ventilated preterm infants were recorded daily during the first 14 days of life to assess the effect on respiratory efforts of recovery from acute respiratory distress syndrome (RDS). Active expiration and persistent asynchrony only occurred during acute RDS (P c 0.01). Throughout the 1Cday period, in the majority of infants making respiratory efforts, a ventilator rate could be found from a standard sequence 30,60,90, 120 breathsjmin which provoked a synchronous interaction, but with increasing postnatal age apnoea became more common (P < 0.01). We conclude that the preterm infants’ spontaneous respiratory efforts are a less important influence on the outcome of mechanical ventilation following recovery from acute RDS. respiratory interaction; mechanical ventilation; respiratory distress syndrome.
Introduction
Preterm neonates frequently make respiratory efforts during mechanical ventilation which can be divided into distinct interactions [5]. The type of respiratory interaction provoked and hence the outcome of mechanical ventilation [5] appears to be dependent on the infant’s spontaneous respiratory timing. Preterm neonates breathe Correspondence CO:Dr A. Greenough, Dept. of Child Health, King’s College Hospital, London SE5 9RS.
U.K. 0378-3782/91/$03.50 0 1991 Elsevier Scientific Publishers Ireland Ltd. Published and Printed in Ireland
70
at rapid respiratory rates, with short inspiratory and expiratory times [4] and high frequency positive pressure ventilation (HFPPV, rates equal to or in excess of 60 breaths/min) closely mimics this inducing synchrony which improves oxygenation [8]. Lengthening inflation time, however, provokes active expiration [ 1,2,6], increasing the risk of pneumothorax [lo]. As a consequence, during acute respiratory distress syndrome (RDS) HFPPV improves outcome compared to ventilation employing rates less than or equal to 40 breathsimin [3,11]. The studies [ 1,2,4-6,8] which have examined the infant’s spontaneous interaction with mechanical ventilation and their respiratory timing have all been carried out in the first week of life and amongst infants suffering from acute respiratory distress. Increasing survival of VLBW infants [9] has meant that many infants require mechanical ventilation even after recovery from RDS and they remain ventilated at least into the second week of life. It cannot be concluded that HFPPV would necessarily be the optimum form of ventilation in these older infants. As RDS resolves, lung compliance may improve which would lengthen the time constant of the respiratory system. Such a change in lung function could result in lengthened inspiratory and expiratory times and thus influence the spontaneous respiratory interactions provoked at different ventilator rates. The aim of this study was to determine if recovery from RDS did indeed influence the nature of the infant’s respiratory efforts during positive mechanical ventilation. Patients and Methods
Consecutive VLBW infants were entered into the study if they required mechanical ventilation at rates > 30 breathsjmin for at least 14 days from birth. Daily measurements were made in two study periods, days U representing acute RDS and from days 5 to 14 representing non-acute respiratory distress. Respiratory measurements were made using the system previously described in detail [4]. Briefly, a pneumotachograph was inserted between the endotracheal tube and ventilator circuit. This recorded flow changes, both due to the infant and the ventilator. This signal was electronically integrated to give volume. Ventilator pressure changes were recorded proximal to the pneumotachograph. Oesophageal pressure changes were recorded using an oesophageal balloon. Flow, volume, oesophageal and ventilator pressure changes were recorded simultaneously on a Gould polygraph. Measurements were made for 20-min periods and only when the infant was stable with blood gases within the clinically acceptable range (pH 7.25-7.4; Pco2 35-45 mmHg; PO* 45-70 mmHg). The infant’s respiratory interaction was determined as that which occurred on more than 80% of positive pressure inflations. The infant’s respiratory interactions were classified as previously defined [7]: (1) apnoea; (2) synchrony; (3) asynchrony. At the initial study the infant’s respiratory interaction was studied at a series of ventilator rates (30, 60, 90 and 120/min). During subsequent study periods infants were studied only at the ventilator rate they were currently receiving but re-studied at a series of rates if they were found to be asynchronous or paralysis had recently been stopped. In the study periods when infants were ventilated at a series of rates
71
they were described as synchronous if synchrony was provoked by at least one of the ventilator rates and subsequently ventilated at the slowest rate at which the synchronous interaction was achieved [6]. Infants were described as asynchronous if they remained asynchronous at all the ventilator rate sequences, they were then paralysed and paralysis continued until the infant’s respiratory distress had improved so that peak pressures could be reduced to 18 cmH20 and rates reduced below 40 breaths/min [lo]. Apnoeic infants were ventilated at the slowest ventilator rate in the sequence [6]. Twenty-seven infants were enrolled into the study with a median gestational age of 27 weeks (range 23-31) and birthweight of 778 g (range 5061500). All of the infants had been ventilated from birth for RDS. Respiratory support was provided by Sechrist ventilators via oral endotracheal tubes. Ethical permission for this study was granted by the King’s College Hospital Ethics Committee. Statistical
analysis
Differences in the respiratory interaction provoked at different postnatal ages were assessed for statistical significance using Fisher’s exact test. Results
During the first four days of life the majority of infants were making respiratory efforts, only 4 of 27 being apnoeic on day 1 (Table I). With increasing postnatal age the proportion of infants without respiratory efforts increased, being greater than 50% from day 7 onwards (day l-4 compared to day 1l-14, P < 0.01). Persistent asynchrony was only seen in the first two days of life, the difference between the
TABLE
I
Spontaneous
Day
respiratory
interactions
Apnoea
in ventilated
infants
(n = 27)
Synchrony
Asynchrony
I 2
4 4
15 II
8” 4”
3 4 5 6 7 8 9 10
4 4 9 11 14 14 14 14
II I1 8 9 IO 10 IO II
II 12 13 14
15 15 15 15
II I1 II II
0 0 0 0 0 0 0 0 0 0 0 0
Paralysed 8 8+4 12 I2 10 7 3 3 3 2 I I I
1
“On day I, 8 infants were asynchronous despite rate manipulation and on day 2, 4 additional became persistently asynchronous, these infants were paralysed on day 1 or 2, respectively.
infants
72
occurrence of apnoea at that time and the rest of the period was highly significant (P < 0.001). All 12 infants persistently asynchronous, despite being studied at the rate sequence, were paralysed. Once pancuronium was discontinued, none of these 12 infants were asynchronous despite being studied at the whole rate sequence, 9 of the infants were apnoeic and 3 had synchronous respiratory efforts. Throughout the 14-day study period, in the majority of infants making respiratory efforts, synchrony could be achieved by increasing the ventilator rate to 60 breaths/min or beyond (Table I). Discussion
In this study we have demonstrated that the majority of infants, even including those born very preterm, were making respiratory efforts during the acute state of their illness. We had previously found [5] that small, very sick premature infants rarely make spontaneous respiratory efforts. In that study [5], however, infants were only studied at one ventilator rate and in the present trial infants were studied on day one at a series of rates; this [8] seems likely to explain the difference in the results of the two studies. Our results have confirmed [3,5] that respiratory interaction usually remains stable at a constant ventilator rate. Except in the first two days of life, infants who were ventilated at a rate which provoked synchrony, remained synchronous (the majority) or became apnoeic, until ventilator rate was altered to be reduced below 30 breaths/min. Our results demonstrate that with increasing postnatal age there is a change in the nature of the preterm infant’s respiratory efforts, asynchrony disappearing and apnoea becoming more common (Table I). These changes may be due to recovery from acute RDS. Infants who are asynchronous have previously been demonstrated to have the most severe RDS [2]. Yet, even in this group, their respiratory status had recovered sufficiently during the second study period (days 5-14) for their ventilator rates and pressure to have been reduced sufficiently so that treatment with pancuronium could be discontinued, indicating recovery from acute RDS. An alternative explanation for the increasing proportion of infants who were apnoeic at the higher postnatal ages is an effect of pancuronium. Although certain of the apnoeic infants had breathed synchronously with the ventilator prior to day 5, the majority had been paralysed during the acute stage of their illness. Once recovery from RDS had commenced, asynchrony could not be provoked despite studying infants at the whole rate sequence. Asynchrony is provoked at low ventilator rates amongst infants with the stiffest lungs [2]. The failure to provoke asynchrony once recovery from RDS had commenced is thus likely to be due to the improvement in lung function occurring at this time. The lack of asynchrony, even at long inspiratory times, after day 4 of life, may explain the low incidence of pneumothorax during recovery from acute RDS [ 141. This study demonstrates that in the second week of life, it is not necessary to manipulate ventilator rate to avoid active expiration [ 1,4,6]. We must therefore conclude that spontaneous respiration seems less likely to cause an airleak during
73
mechanical ventilation in the second rather than the first week of life [5,10]. Although synchrony could be provoked at a standard fast ventilator rate in infants who made respiratory efforts in the second week of life, this interaction has not been proved to improve oxygenation following recovery from RDS. Increasing numbers of infants ventilated beyond the first week of life now make it imperative to determine, by blood gas measurements at a series of ventilator rates, optimum settings for such infants. References 1 2 3 4 5 6 I 8
9 10
I1 12
13 14
Field, D., Milner, A.D. and Hopkin, I.E. (1984): High and conventional rates of positive pressure ventilation. Arch. Dis. Child., 59, 1151-1154. Greenough, A., (1988): The premature infant’s respiratory response to mechanical ventilation. Early Hum. Dev., 17, l-5. Greenough, A., Dixon, A. and Roberton, N.R.C. (1984): Pulmonary interstitial emphysema. Arch. Dis. Child., 59, 1046-l 105. Greenough, A., Greenall, F. and Gamsu, H. (1987): Synchronous respiration - which ventilator rate is best? Acta. Paediatr. &and., 76, 713-718. Greenough, A., Morley, C.J. and Davis, J.A. (1983): Interaction of spontaneous respiration and artificial ventilation. J. Pediatr., 103, 769-773. Greenough, A., Morley, C.J. and Pool, J. (1986): Fighting the ventilator - are fast rates an effective alternative to paralysis? Early Hum. Dev., 13, 189-194. Greenough, A. and Greenall, F. (1988): Observation of spontaneous respiratory interaction with artificial ventilation. Arch. Dis. Child., 63, 168-171. Greenough, A., Pool, J., Greenall, F., Morley, C.J. and Gamsu, H. (1987): Comparison of different rates of artificial ventilation in preterm neonates with the respiratory distress syndrome. Acta. Paediatr. &and., 76, 70&12. Greenough, A. and Roberton, N.R.C. (1985): Morbidity and mortality in neonates ventilated for the respiratory distress syndrome. Br. Med. J., 290, 597-600. Greenough, A., Wood, S., Morley, C.J. and Davis, J.A. (1984): Pancuronium prevents pneumothoraces in ventilated premature infants who actively expire against positive pressure ventilation. Lancet, i, l-3. Morley, C.J. and Greenough, A. (1990): Respiratory compliance measurements in very premature babies treated with artificial surfactant (ALEC). Arch. Dis. Child., in press. Heicher, D.A., Kastings, D.S. and Richards, J.R. (1981): Prospective clinical comparison of two methods for mechanical ventilation of neonates: rapid rate and short expiratory time. J. Pediatr., 98, 957-961. Hird, M., Greenough, A. and Gamsu, H. (1990) Gas trapping during high frequency pressure ventilation using conventional ventilators. Early Hum. Dev., 22, 51-56. Tarnow-Mordi, W.O., Narang, A. and Wilkinson, A.R. (1985): Lack of association barotrauma and airleak in hyaline membrane disease. Arch. Dis. Child., 60, 555-559.
positive between
