Downloaded from http://fn.bmj.com/ on May 26, 2015 - Published by group.bmj.com
Short research report
Cerebral oxygenation with different nasal continuous positive airway pressure levels in preterm infants Stefano Bembich, Laura Travan, Gabriele Cont, Jenny Bua, Tamara Strajn, Sergio Demarini Institute for Maternal and Child Health–IRCCS “Burlo Garofolo” – Trieste, Italy Correspondence to Dr Sergio Demarini, Division of Neonatology, Institute for Maternal and Child Health, IRCCS “Burlo Garofolo” Via dell’Istria, 65, Trieste I-34137, Italy; [email protected] Received 5 March 2014 Revised 26 September 2014 Accepted 1 October 2014 Published Online First 21 October 2014
ABSTRACT Objectives This study evaluates the effect of varying nasal continuous positive airway pressure (NCPAP) level on cerebral blood flow (CBF) and oxygenation in preterm infants. Methods Oxy-haemoglobin (HbO2) and total haemoglobin (HbTot), as CBF estimates, and the ratio between HbO2 and HbTot (HbO2/HbTot), as cerebral oxygenation estimate, were assessed by near-infrared spectroscopy in 26 stable preterm newborns at a postmenstrual age between 26 and 33 weeks. Baseline HbO2, HbTot and HbO2/HbTot values were initially collected with NCPAP at 5 cm H2O and then compared with values obtained with NCPAP levels at both 3 and 8 cm H2O. Results Compared with 5 cm H2O, cerebral HbO2, HbTot and HbO2/HbTot remained unchanged both after increasing (to 8 cm H2O) and decreasing (to 3 cm H2O) the NCPAP level. This result was observed both in regional areas (24 sites) and in the overall monitored area (frontal and parietal cortex). Compared with 8 cm H2O, peripheral oxygen saturation significantly decreased at 3 cm H2O ( p=0.021). Heart rate did not change. Conclusions No differences in CBF and cerebral oxygenation were observed with NCPAP levels in the range 3–8 cm H2O despite a decrease in peripheral oxygenation with 3 cm H2O.
3–8 cm H2O on cerebral blood flow (CBF) and oxygenation in preterm infants. We tested the hypothesis that, within the range studied, different NCPAP levels would not affect cerebral oxygenation.
MATERIALS AND METHODS Participants The study was performed in the Neonatal Intensive Care Unit of the Institute for Maternal and Child Health, IRCCS ‘Burlo Garofolo’ (Trieste, Italy). We enrolled 26 stable Caucasian preterm newborns admitted to our unit between 1 April and 31 December 2013. They were all treated with NCPAP at a pressure level of 5 cm H2O. Table 1 shows clinical and demographic data. The fraction of inspired oxygen was maintained between 0.21 and 0.25, and it was adjusted to keep peripheral oxygen saturation (SpO2) in a range between 91% and 95%.7 Cerebral lesions were excluded by ultrasound. Infants were on caffeine and vitamin D. The Independent Committee for Bioethics of our hospital approved the research, and informed consent was obtained from parents.
NIRS recording INTRODUCTION
To cite: Bembich S, Travan L, Cont G, et al. Arch Dis Child Fetal Neonatal Ed 2015;100:F165–F168.
Nasal continuous positive airway pressure (NCPAP) is commonly used in preterm infants with respiratory distress. The optimal NCPAP pressure is unknown, although pressures of 4–6 cm H2O are traditionally used.1 2 In adults, with increasing pressure levels, cardiac output decreases3 due to increased intrathoracic pressure. Although this was not observed in preterm infants,4 adverse effects on systemic and pulmonary venous return have been reported.5 More recently, extubation failure in preterm infants was found to be lower with NCPAP pressures of 7–9 than with 4–6 cm H2O.2 Using one-channel near-infrared spectroscopy (NIRS), Dani et al6 found that NCPAP levels between 2 and 6 cm H2O affected neither cerebral oxygenation nor cerebral blood volume. However, the NCPAP pressure range was between 2 and 6 cm H2O. Thus, the effect of higher and potentially beneficial NCPAP levels on cerebral haemodynamics was not examined. The aim of our study was to evaluate, by NIRS, the effects of varying NCPAP levels in the range of
CBF and oxygenation were monitored by a multichannel NIRS device (ETG-100, Hitachi, Tokyo, Japan). It monitors changes in oxy-haemoglobin (HbO2) and deoxy-haemoglobin (Hbb) concentration from 24 sites (channels) on the cortex by 16 optical fibres of 1 mm in diameter (optodes) placed on the scalp (eight light emitters and eight light detectors). Near-infrared light is emitted at two wavelengths, 780 and 830 nm, through the fibres, and its intensity is modulated at different frequencies among emitters, ranging from 1 to 6.5 kHz. The reflected light is then sampled at a frequency of 10 Hz. Variation in total haemoglobin (HbTot) concentration is derived by summing changes in HbO2 and Hbb. The device measures haemoglobin changes in mM×mm, which is the product of the haemoglobin concentration expressed in millimolar and the optical path length expressed in millimetres. The optodes were arranged in a 4×4 pattern by a fibre holder, with an inter-optodic distance of 2.5 cm, and were positioned above the frontal and parietal cortex of the neonate (figure 1).8
Bembich S, et al. Arch Dis Child Fetal Neonatal Ed 2015;100:F165–F168. doi:10.1136/archdischild-2014-306356
F165
Downloaded from http://fn.bmj.com/ on May 26, 2015 - Published by group.bmj.com
Short research report Table 1 Clinical and demographic data Sex Weight at birth (g) Weight at observation (g) Gestational age at birth (weeks) Postmenstrual age at observation (weeks) Age at observation (days)
11 females, 15 males 1325 (540–2050) 1222 (709–2070) 30 (25–32) 31 (26–33) 6 (2–21)
Data are expressed as median (range).
Procedure All the infants were tested in supine position. NCPAP was provided by the Leoni plus ventilator (Heinen &Löwerstein, Bad Ems, Germany). A 60 s baseline was collected, with NCPAP level at 5 cm H2O. Then, NCPAP pressure was changed to 3 cm H2O and NIRS data were recorded for 10 min. Subsequently, NCPAP level was put back to 5 cm H2O for additional 10 min. Then, the same procedure was repeated changing NCPAP pressure to 8 cm H2O. The ventilator we used keeps the level of NCPAP constant. The sequence of NCPAP level variation (e.g. first at 3 cm H2O and then at 8 cm H2O, or vice versa) was equally distributed among participants by randomisation (binary number random generation). Heart rate and peripheral oxygen saturation (SpO2) were continuously monitored during the procedure.
Data analysis Our analysis focused on variation of HbO2 and HbTot, as estimate of CBF,9 10 and of the ratio between HbO2 and HbTot (HbO2/HbTot), as estimate of cerebral oxygenation.11 Signal components possibly related to physiological noise or movement were removed by filtering the NIR signal.8
Figure 1 Representation of optical fibres position on a schematic newborn’s head. Red dots indicate near-infrared light emitters, and blue dots indicate near-infrared light detectors. The 24 detecting channels are indicated by squares. Reference points we take into account to have a reliable positioning of fibre holder are also evidenced (nasion/inion and Cz/T3/T4 points of the international 10–20 EEG electrode placement system). F166
Since we had no baseline data on which to calculate a sample size, we commenced our study without a sample size calculation. After collecting data from 16 babies, we observed a mean difference in HbO2 of 0.088 mM×mm in the overall monitored area after decreasing NCPAP from 5 cm H2O (mean=−0.067 mM×mm; SD=0.110 mM×mm) to 3 cm H2O (mean=−0.155 mM×mm; SD=0.134 mM×mm). Based on these data, we calculated that a sample size of at least 24 infants would be required to detect a significant variation in HbO2 in the overall monitored area, after decreasing NCPAP level from 5 to 3 cm H2O, with 80% power and a 0.05 significance criterion. The device we used is a continuous-wave NIRS system that does not allow absolute measurements of cerebral oxygenation and haemodynamics.12 Therefore, results are based on comparisons between discrete relative measurements. Data were analysed taking into account both single channels separately (regional variations) and the overall monitored area. For every channel, a baseline at 5 cm H2O level was calculated as the mean of cortical HbO2, HbTot and HbO2/HbTot concentration changes, in the 60 s preceding NCPAP pressure change, by a program written in Matlab format (MathWorks, Natick, Massachusetts). Variations in CBF and in oxygenation were estimated as the mean concentration changes in, respectively, HbO2, HbTot and HbO2/HbTot during the 10th minute following NCPAP increase or decrease. We considered this specific and discrete period of detection as the most appropriate to extract a continuous and stable signal, with a sufficient delay from the NCPAP variation. A longer period of time detection, e.g. 15 or 20 min, would have been associated with relevant newborn’s movement noise, frequently due to an increasing intolerance to the fibre holder. By two-tailed paired t tests, we compared, for every channel, HbO2, HbTot and HbO2/HbTot mean concentration changes during baseline and during the 10th minute after NCPAP variation. We used a false discovery rate approach to control type I error in multiple testing situations (q=0.05).13 Cortical blood flow and oxygenation in the overall monitored area were studied by averaging HbO2, HbTot and HbO2/HbTot concentration changes among all 24 channels during baseline and during the 10th minute after the NCPAP variation. Then, the overall HbO2, HbTot and HbO2/HbTot mean changes during baseline (5 cm H2O) and after NCPAP increase (8 cm H2O) or decrease (3 cm H2O) were compared by repeated measures analysis of variance (ANOVA), with post hoc comparisons to baseline of both NCPAP variations by two-tailed paired t test. Heart rate and SpO2 changes, after increasing to 8 cm H2O or decreasing to 3 cm H2O the NCPAP level, were analysed by paired t-test as well. Heart rate and SpO2 variation after NCPAP level change were estimated, by the instrument software (Intellivue MP 20, Philips, Amsterdam, The Netherlands), as the mean change during the 10th minute following NCPAP variation to 8 or to 3 cm H2O. By two-tailed paired t-test, mean heart rate and SpO2 observed during the 10th minute after NCPAP increase to 8 cm H2O were compared with those observed during the 10th minute after NCPAP decrease to 3 cm H2O.
RESULTS No channel showed a significant difference (p
Short research report
Cerebral oxygenation with different nasal continuous positive airway pressure levels in preterm infants Stefano Bembich, Laura Travan, Gabriele Cont, Jenny Bua, Tamara Strajn, Sergio Demarini Institute for Maternal and Child Health–IRCCS “Burlo Garofolo” – Trieste, Italy Correspondence to Dr Sergio Demarini, Division of Neonatology, Institute for Maternal and Child Health, IRCCS “Burlo Garofolo” Via dell’Istria, 65, Trieste I-34137, Italy; [email protected] Received 5 March 2014 Revised 26 September 2014 Accepted 1 October 2014 Published Online First 21 October 2014
ABSTRACT Objectives This study evaluates the effect of varying nasal continuous positive airway pressure (NCPAP) level on cerebral blood flow (CBF) and oxygenation in preterm infants. Methods Oxy-haemoglobin (HbO2) and total haemoglobin (HbTot), as CBF estimates, and the ratio between HbO2 and HbTot (HbO2/HbTot), as cerebral oxygenation estimate, were assessed by near-infrared spectroscopy in 26 stable preterm newborns at a postmenstrual age between 26 and 33 weeks. Baseline HbO2, HbTot and HbO2/HbTot values were initially collected with NCPAP at 5 cm H2O and then compared with values obtained with NCPAP levels at both 3 and 8 cm H2O. Results Compared with 5 cm H2O, cerebral HbO2, HbTot and HbO2/HbTot remained unchanged both after increasing (to 8 cm H2O) and decreasing (to 3 cm H2O) the NCPAP level. This result was observed both in regional areas (24 sites) and in the overall monitored area (frontal and parietal cortex). Compared with 8 cm H2O, peripheral oxygen saturation significantly decreased at 3 cm H2O ( p=0.021). Heart rate did not change. Conclusions No differences in CBF and cerebral oxygenation were observed with NCPAP levels in the range 3–8 cm H2O despite a decrease in peripheral oxygenation with 3 cm H2O.
3–8 cm H2O on cerebral blood flow (CBF) and oxygenation in preterm infants. We tested the hypothesis that, within the range studied, different NCPAP levels would not affect cerebral oxygenation.
MATERIALS AND METHODS Participants The study was performed in the Neonatal Intensive Care Unit of the Institute for Maternal and Child Health, IRCCS ‘Burlo Garofolo’ (Trieste, Italy). We enrolled 26 stable Caucasian preterm newborns admitted to our unit between 1 April and 31 December 2013. They were all treated with NCPAP at a pressure level of 5 cm H2O. Table 1 shows clinical and demographic data. The fraction of inspired oxygen was maintained between 0.21 and 0.25, and it was adjusted to keep peripheral oxygen saturation (SpO2) in a range between 91% and 95%.7 Cerebral lesions were excluded by ultrasound. Infants were on caffeine and vitamin D. The Independent Committee for Bioethics of our hospital approved the research, and informed consent was obtained from parents.
NIRS recording INTRODUCTION
To cite: Bembich S, Travan L, Cont G, et al. Arch Dis Child Fetal Neonatal Ed 2015;100:F165–F168.
Nasal continuous positive airway pressure (NCPAP) is commonly used in preterm infants with respiratory distress. The optimal NCPAP pressure is unknown, although pressures of 4–6 cm H2O are traditionally used.1 2 In adults, with increasing pressure levels, cardiac output decreases3 due to increased intrathoracic pressure. Although this was not observed in preterm infants,4 adverse effects on systemic and pulmonary venous return have been reported.5 More recently, extubation failure in preterm infants was found to be lower with NCPAP pressures of 7–9 than with 4–6 cm H2O.2 Using one-channel near-infrared spectroscopy (NIRS), Dani et al6 found that NCPAP levels between 2 and 6 cm H2O affected neither cerebral oxygenation nor cerebral blood volume. However, the NCPAP pressure range was between 2 and 6 cm H2O. Thus, the effect of higher and potentially beneficial NCPAP levels on cerebral haemodynamics was not examined. The aim of our study was to evaluate, by NIRS, the effects of varying NCPAP levels in the range of
CBF and oxygenation were monitored by a multichannel NIRS device (ETG-100, Hitachi, Tokyo, Japan). It monitors changes in oxy-haemoglobin (HbO2) and deoxy-haemoglobin (Hbb) concentration from 24 sites (channels) on the cortex by 16 optical fibres of 1 mm in diameter (optodes) placed on the scalp (eight light emitters and eight light detectors). Near-infrared light is emitted at two wavelengths, 780 and 830 nm, through the fibres, and its intensity is modulated at different frequencies among emitters, ranging from 1 to 6.5 kHz. The reflected light is then sampled at a frequency of 10 Hz. Variation in total haemoglobin (HbTot) concentration is derived by summing changes in HbO2 and Hbb. The device measures haemoglobin changes in mM×mm, which is the product of the haemoglobin concentration expressed in millimolar and the optical path length expressed in millimetres. The optodes were arranged in a 4×4 pattern by a fibre holder, with an inter-optodic distance of 2.5 cm, and were positioned above the frontal and parietal cortex of the neonate (figure 1).8
Bembich S, et al. Arch Dis Child Fetal Neonatal Ed 2015;100:F165–F168. doi:10.1136/archdischild-2014-306356
F165
Downloaded from http://fn.bmj.com/ on May 26, 2015 - Published by group.bmj.com
Short research report Table 1 Clinical and demographic data Sex Weight at birth (g) Weight at observation (g) Gestational age at birth (weeks) Postmenstrual age at observation (weeks) Age at observation (days)
11 females, 15 males 1325 (540–2050) 1222 (709–2070) 30 (25–32) 31 (26–33) 6 (2–21)
Data are expressed as median (range).
Procedure All the infants were tested in supine position. NCPAP was provided by the Leoni plus ventilator (Heinen &Löwerstein, Bad Ems, Germany). A 60 s baseline was collected, with NCPAP level at 5 cm H2O. Then, NCPAP pressure was changed to 3 cm H2O and NIRS data were recorded for 10 min. Subsequently, NCPAP level was put back to 5 cm H2O for additional 10 min. Then, the same procedure was repeated changing NCPAP pressure to 8 cm H2O. The ventilator we used keeps the level of NCPAP constant. The sequence of NCPAP level variation (e.g. first at 3 cm H2O and then at 8 cm H2O, or vice versa) was equally distributed among participants by randomisation (binary number random generation). Heart rate and peripheral oxygen saturation (SpO2) were continuously monitored during the procedure.
Data analysis Our analysis focused on variation of HbO2 and HbTot, as estimate of CBF,9 10 and of the ratio between HbO2 and HbTot (HbO2/HbTot), as estimate of cerebral oxygenation.11 Signal components possibly related to physiological noise or movement were removed by filtering the NIR signal.8
Figure 1 Representation of optical fibres position on a schematic newborn’s head. Red dots indicate near-infrared light emitters, and blue dots indicate near-infrared light detectors. The 24 detecting channels are indicated by squares. Reference points we take into account to have a reliable positioning of fibre holder are also evidenced (nasion/inion and Cz/T3/T4 points of the international 10–20 EEG electrode placement system). F166
Since we had no baseline data on which to calculate a sample size, we commenced our study without a sample size calculation. After collecting data from 16 babies, we observed a mean difference in HbO2 of 0.088 mM×mm in the overall monitored area after decreasing NCPAP from 5 cm H2O (mean=−0.067 mM×mm; SD=0.110 mM×mm) to 3 cm H2O (mean=−0.155 mM×mm; SD=0.134 mM×mm). Based on these data, we calculated that a sample size of at least 24 infants would be required to detect a significant variation in HbO2 in the overall monitored area, after decreasing NCPAP level from 5 to 3 cm H2O, with 80% power and a 0.05 significance criterion. The device we used is a continuous-wave NIRS system that does not allow absolute measurements of cerebral oxygenation and haemodynamics.12 Therefore, results are based on comparisons between discrete relative measurements. Data were analysed taking into account both single channels separately (regional variations) and the overall monitored area. For every channel, a baseline at 5 cm H2O level was calculated as the mean of cortical HbO2, HbTot and HbO2/HbTot concentration changes, in the 60 s preceding NCPAP pressure change, by a program written in Matlab format (MathWorks, Natick, Massachusetts). Variations in CBF and in oxygenation were estimated as the mean concentration changes in, respectively, HbO2, HbTot and HbO2/HbTot during the 10th minute following NCPAP increase or decrease. We considered this specific and discrete period of detection as the most appropriate to extract a continuous and stable signal, with a sufficient delay from the NCPAP variation. A longer period of time detection, e.g. 15 or 20 min, would have been associated with relevant newborn’s movement noise, frequently due to an increasing intolerance to the fibre holder. By two-tailed paired t tests, we compared, for every channel, HbO2, HbTot and HbO2/HbTot mean concentration changes during baseline and during the 10th minute after NCPAP variation. We used a false discovery rate approach to control type I error in multiple testing situations (q=0.05).13 Cortical blood flow and oxygenation in the overall monitored area were studied by averaging HbO2, HbTot and HbO2/HbTot concentration changes among all 24 channels during baseline and during the 10th minute after the NCPAP variation. Then, the overall HbO2, HbTot and HbO2/HbTot mean changes during baseline (5 cm H2O) and after NCPAP increase (8 cm H2O) or decrease (3 cm H2O) were compared by repeated measures analysis of variance (ANOVA), with post hoc comparisons to baseline of both NCPAP variations by two-tailed paired t test. Heart rate and SpO2 changes, after increasing to 8 cm H2O or decreasing to 3 cm H2O the NCPAP level, were analysed by paired t-test as well. Heart rate and SpO2 variation after NCPAP level change were estimated, by the instrument software (Intellivue MP 20, Philips, Amsterdam, The Netherlands), as the mean change during the 10th minute following NCPAP variation to 8 or to 3 cm H2O. By two-tailed paired t-test, mean heart rate and SpO2 observed during the 10th minute after NCPAP increase to 8 cm H2O were compared with those observed during the 10th minute after NCPAP decrease to 3 cm H2O.
RESULTS No channel showed a significant difference (p
