Acta Padiatr 81: 389-93. 1992

Changes in cerebral oxygenation and cerebral blood volume during endotracheal suctioning in ventilated neonates L Skov, J Ryding, 0 Pryds and G Greisen Department of Neonatology. Rigshospitalet. Copenhagen. Denmark

Skov L, Ryding J, Pryds 0, Greisen G. Changes in cerebral oxygenation and cerebral blood volume during endotracheal suctioning in ventilated neonates. Acta Paediatr 1992;81:389-93. Stockholm. ISSN 0803-5253 The effect of endotracheal suctioning on cerebral haemodynamics was investigated in 29 newborn infants with a mean gestational age of 31 weeks (range 25-40 weeks). Prior to one of two suctioning procedures, the inspiratory fraction of oxygen was increased by 10%. Brain oxygenation and total haemoglobin concentration were estimated continuously by near infrared spectroscopy.Mean arterial blood pressure, arterial blood oxygen saturation and carbon dioxide tension were recorded simultaneously. Brain oxygenation decreased in parallel with arterial oxygen saturation during suctioning. Preoxygenation ameliorated the decrease in brain oxygenation and arterial oxygen saturation whereas there was no benefit with regard to the changes in total haemoglobin concentration, carbon dioxide tension or mean arterial pressure. Changes in total haemoglobin concentration were related closely to concomitant changes in carbon dioxide tension ( p < 0.0001) but unrelated to changes in mean arterial pressure or arterial oxygen saturation. Our findings suggest that cerebral blood volume may react to changes in carbon dioxide tension during endotracheal suctioning in mechanically ventilated neonates. Apparently, preoxygenation prior to suctioning does not ameliorate the stress in normoxic infants. 0 Cerebral haemodynamics. endotracheal suctioning, near infrared spectroscopy L Skoo, Department of Neonatology, Rigshospitalet. Blegdamsuej 9 , 2100 Copenhagen, Denmark

Endotracheal suctioning of mechanically ventilated, newborn infants may induce slight hypoxia and changes in arterial blood pressure. These events could have adverse effects on the cerebral haemodynamic system of sick neonates if vasoreactivity is impaired. Several procedures have been proposed to minimize the physiological responses during suctioning, for example by using preoxygenation or manual ventilation prior to the episode (1). The evaluation of ameliorating care procedures has been based on direct blood pressure recordings (2), on changes in arterial oxygen saturation or oxygen tensions (3,4), and on changes in cerebral blood flow velocity (5). Near infrared spectrophotometry (NIRS) allows noninvasive monitoring of changes in cerebral blood volume and in brain oxygenation. The purposes of this study were: to evaluate the effect of endotracheal suctioning on cerebral oxygenation and haemodynamics estimated by means of NIRS; to investigate whether induced changes in cerebral haemodynamics were related to changes in arterial blood pressure or blood gases; and to investigate whether preoxygenation was beneficial in terms of reducing the fluctuation in physiological parameters.

Patients and methods Twenty-nine newborn infants with a mean gestational

age of 3 1 weeks (range 25-40 weeks) and a mean birth weight of 1770 g (range 690-3775 g) were investigated. The Apgar score at 5 min and umbilical cord pH averaged 8 (range 5-10) and 7.2 (range 6.7-7.4), respectively. Twenty-four preterm infants were mechanically ventilated because of respiratory distress syndrome, whereas five term infants were assisted because of neonatal asphyxia. The inspiratory fraction of oxygen ranged from 0.21 to 0.90. According to our care regimens, all infants were sedated with phenobarbital (loading dose 15 mg/kg; maintenance dose 5 mg/kg) and were paralysed with pancuronium (0.10 mg/kg). Mean postnatal age was 36 h (range 14-94 h). Five patients had intraventricular haemorrhage grade 1-11 and four grade 111-IV (6). Changes in cerebral haemoglobin concentration (tHb), oxygenated and deoxygenated haemoglobin (HbOz and HB, respectively) were measured by NIRS, as described previously (7). NIRS depends on the relative transparency of the neonate’s head to near infrared light, which is attenuated by scattering and absorption in the tissues (8). If the amount of scattered light is assumed to be constant, changes in optical attenuation depend only on the concentration of lightabsorbing chromophores and whether they are in the oxygenated or deoxygenated forms. This assumption is based on the Lambert-Beer’s law which states that a proportional relationship exists between the optical attenuation and the concentration of the chromophore,

390

L Skor et al.

ACTA PRDIATR 81 (1992)

mM x cm 0.1

..........

,.-*.-.4..-.

..............

\.*,."

,.

I\---

.-..-

.

.-.--.-.*.I-'".-.-

---..-.-,,1e

I

0

-.r.-.-

.,.-._

........................

-'*---.-"\,-v

1

[

[:

mmHg

kPa

20 rnin

Fig. 1. Changes in tHb, 0 1 , SaOz, PaC02 and MABP during endotracheal suctioning of a newborn infant.

the extinction coefficient of the chromophore and the path length of the light. By selection of appropriate wavelengths, algorithms can be developed for the calculation of changes in the concentration of oxygenated and deoxygenated haemoglobin, respectively (9). The NIR instrument (Radiometer, Denmark) uses four semiconductor laser diodes with wavelengths from 775 nm to 904 nm. The lasers are operated sequentially and pulsed for 200 ns with a repetition rate of 500 Hz. The light is transferred by means of an optical cable, ending in an optode which is fixed perpendicular to the infant's head. An identical optode placed on the opposite side of the head collects and conducts back the received photons to another semiconductor diode, converting light into an electrical signal. The controlling computer calculates the amount of light attenuation within the head and converts the value to changes in HbO2 and HB by means of the algorithm. Thus, total cerebral haemoglobin concentration (tHb) is the sum of HbO2 and HB whereas the haemoglobin oxygenation index (01) is the difference between HbO2 and HB (Fig. 1). In the present study the NIRS recordings were performed in transmission mode in low-birth-weight infants, whereas larger infants were investigated in reflection mode. In both modes, the relation between optical path length and inter-optode spacing appears to be constant as observed recently in time of flight studies (10). In all patients, the inter-optode spacing was more than 5 cm but because the path length is unknown, the present results are relative and are expressed as mM x cm.

Mean arterial blood pressure (MABP) was recorded from indwelling arterial catheters; carbon dioxide tension (PaC02) was monitored by means of a calibrated transcutaneous electrode (TCM3, Radiometer, Denmark) and arterial blood oxygenation saturation (SaOz) was measured by pulse oximetry (Ohmeda) from the right hand. Analogue data were transferred to a computer (Hewlett-Packard Vectra 05/20) and stored for later analysis. Experimental design All infants had been undisturbed for at least 30 min prior to the study. Within approximately 2 h, routine endotracheal suctioning (lasting 10 to 15 s) was performed twice by the attending nurse; suctioning was performed with and without preoxygenation, respectively. In 15 infants preoxygenation was instituted prior to the first suctioning procedure and in the remaining cases prior to the second episode. Preoxygenation was achieved by increasing the inspiratory oxygen fraction by approximately 10% for a 5-min period. Calculations Data from three episodes during each suctioning procedure were selected for analysis: 1 min before the suctioning (baseline), when the 01-signal deflection was maximal and 1 min after completion of the suctioning procedure. The results were normalized according to differences from baseline values. Because the transcuta-

Cerebral haemodynamics during endotracheal suctioning

ACTA PRDIATR 81 (1992)

neous C02 electrode may be delayed, all data. were reanalysed after adjusting for an estimated time lag of 20 s. Statistics The null hypothesis to be tested was that 0 1 and tHb remained constant during the endotracheal suctioning. Intra-individual changes in 01, tHb, Sa02, PaC02 and MABP were analysed by means of analysis of variance (ANOVA), and the student Newman-Keuls test was used for multiple comparisons. The relation between changes in 0 1 and tHb, and changes in Sa02,PaC02and MABP were analysed by univariate or multiple regression. Because of the repeated measures design, a factor with 29 levels was included. Thus, the linear regression analysis tested only the intra-individual changes ( 1 1). The statistical program SPSS (Statistical Package for the Social Sciences, Chicago, Ill.) was used and the level of significance was set at 0.05. The study was approved by the Ethics Committee of Greater Copenhagen and parental informed consent was obtained for each infant.

Results Changes in OI and tHb during endotracheal suctioning During the suctioning procedure a significant decrease in 0 1 was observed (mean -0.15 mM x cm; 95% confidence interval - 0.18 to -0.12 mM x cm) whereas tHb remained constant (mean 0.003 mM x-cm; 95% confidence interval -0.01 1 to 0.017 mM x cm). One minute after suctioning, 0 1 was still below the baseline value (mean -0.09 mM x cm; 95% confidence interval

39 1

-0.13 to -0.05 mM x cm) and tHb had increased slightly (mean 0.01 3 mM x cm; 95% confidence interval -0.001 to 0.027 mM x cm). Changes in SaO2, PaCO2 and MA BP during endotracheal suctioning In all infants endotracheal suctioning resulted in a decrease in SaO2 (mean -7.2%; 95% confidence interval -8.8 to -5.7%), and S a 0 2 did not reach the baseline level 1 min after the procedure. During suctioning PaC02 increased slightly (mean 0.14 kPa; 95% confidence interval 0.06 to 0.22 kPa) and PaCO2 was even higher after 1 min (mean 0.22 kPa; 95% confidence interval 0.14 to 0.3 1 kPa). These values were not affected notably after adjusting for an estimated time lag of 20 s for the transcutaneous COz electrode. MABP tended to be constant although wide inter-individual changes were observed (range -9.8 to 18.4 mmHg). Changes in OI us changes in SaO2, PaCO2 and MABP Changes in 0 1 were related closely to changes in SaO2 ( p < 0.0001) (Fig. 2), whereas there was no relation between changes in 0 1 and changes in PaC02 or MABP, respectively. However, when changes in Sa02 were adjusted for, a negative relation between changes in 0 1 and PaC02 was found ( p < 0.05). Eflect of preoxygenation Preoxygenation diminished the decrease in SaOz and 0 1 during endotracheal suctioning ( p < 0.05). However, changes in tHb (mean -0.01 1 mM x cm; 95% confi-

01 mM.cm

0.1 r

.,. ..... . . I

OI l a -o.3t

. -.., . . ......... . . . .. . . . ... .. ....... ...... . ...... ... .. .. . ..I

I .

-0.1

-

.I

I

0.2

-0.4

t

- 0.5I - 25

.I

.

. I

-20

I

-15

I

I

- 10

-5

I

I

0

5

1

10

Sa02% Fig. 2. The relation between changes in 01 and changes in SaOz during endotracheal suctioning in 29 mechanically ventilated neonates (p