Article Text

Download PDFPDF

Skin-to-skin care alters regional ventilation in stable neonates
  1. Nicholas F Schinckel1,2,
  2. Leah Hickey1,2,3,
  3. Elizabeth J Perkins1,
  4. Prue M Pereira-Fantini1,
  5. Sienna Koeppenkastrop1,2,
  6. Isabella Stafford1,4,
  7. Georgie Dowse1,2,
  8. David G Tingay1,2,3
  1. 1 Neonatal Research, Murdoch Children’s Research Institute, Parkville, Victoria, Australia
  2. 2 Department of Paediatrics, The University of Melbourne, Melbourne, Victoria, Australia
  3. 3 Department of Neonatal Medicine, The Royal Children’s Hospital Melbourne, Parkville, Victoria, Australia
  4. 4 Menzies Institute for Medical Research, University of Tasmania, Hobart, Tasmania, Australia
  1. Correspondence to Nicholas F Schinckel, Neonatal Research, Murdoch Childrens Research Institute, Parkville, Victoria, Australia; nick.schinckel{at}


Objective Skin-to-skin care (SSC) has proven psychological benefits; however, the physiological effects are less clearly defined. Regional ventilation patterns during SSC have not previously been reported. This study aimed to compare regional ventilation indices and other cardiorespiratory parameters during prone SSC with supine and prone position cot-nursing.

Design Prospective observational study.

Setting Single quaternary neonatal intensive care unit in Australia.

Patients 20 infants spontaneously breathing (n=17) or on non-invasive ventilation (n=3), with mean (SD) gestational age at birth of 33 (5) weeks.

Interventions Thirty-minute episodes of care in each position: supine cot care, prone SSC and prone cot care preceding a 10 min period of continuous electrical impedance tomography measurements of regional ventilation.

Main outcome measures In each position, ventral–dorsal and right–left centre of ventilation (CoV), percentage of whole lung ventilation by region and percentage of apparent unventilated lung regions were determined. Heart and respiratory rates, oxygen saturation and axillary temperature were also measured.

Results Heart and respiratory rates, oxygen saturation, temperature and right-left lung ventilation did not differ between the three positions (mixed-effects model). Ventilation generally favoured the dorsal lung, but the mean (95% CI) ventrodorsal CoV was −2.0 (−0.4 to –3.6)% more dorsal during SSC compared with prone. Supine position resulted in 5.0 (1.5 to 5.3)% and 4.5 (3.9 to 5.1)% less apparently unventilated lung regions compared with SSC and prone, respectively.

Conclusions In clinically stable infants, SSC generates a distinct regional ventilation pattern that is independent of prone position and results in greater distribution of ventilation towards the dorsal lung.

  • respiratory
  • neonatology
  • monitoring
  • nursing

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

What is already known on this topic?

  • Skin-to-skin care (SSC) has associated psychological benefits for both infants and parents, including decreased maternal anxiety, reduced infant crying and increased cognitive development .

  • Other clinical benefits include improved survival to discharge, improved weight gain and increase settled sleep.

  • There is uncertainty surrounding the physiological effects of SSC on the infant, importantly the effect on respiration.

What this study adds?

  • This study is the first to observe regional ventilation during SSC and showed an effect on respiratory patterns; in particular, greater dorsal ventilation.

  • This work demonstrates that respiratory patterns may not be universal for all infants and that different populations and pathologies need to be considered.

  • The findings provide further evidence of general clinical stability for the infant during SSC with no observed changes in heart rate, respiratory rate, SpO2 or temperature.


Skin-to-skin care (SSC), wherein a baby is positioned directly onto a reclined caregiver’s chest, has proven benefits for parents and babies, including improvements in attachment behaviour, reduced maternal anxiety,1 2 increased settled sleep,3 reduced crying,4 enhanced breast feeding and improved weight gain.5 Importantly, it is also associated with increased survival to discharge3 6 and better cognitive development at 10 years.7 8 The cardiorespiratory effects of SSC, however, are less clearly defined. Bradycardia, normocardia and tachycardia have been reported,9–13 and either no effect or decreases in respiratory rate (RR) and peripheral oxygen saturation (SpO2) have been observed.10 11 13–15

Previous studies examining the impact of SSC on respiratory outcomes have used non-specific global measures, such as RR and SpO2, neither of which accurately reflect respiratory volumes or mechanics, nor account for the impact of lung development and disease states. In contrast, measures of regional ventilation, such as electrical impedance tomography (EIT), describe the distribution of ventilation within the lung, thereby providing a more accurate measure of alveolar function.16 Regional ventilation has not been described during SSC, but posture and disease and/or developmental state are known to alter regional ventilation in preterm infants.17–19

The aims of this study were to investigate the effects of SSC on regional ventilation and other cardiorespiratory parameters compared with cot-nursing in supine and prone positions. We hypothesised that ventilation would be more homogeneous during SSC and prone position compared with supine. The primary outcome was ventilation homogeneity, assessed using standardised numerical measures of regional ventilation.16 The role of nursing position on heart rate (HR), SpO2, RR, inspiratory time (Ti) and axillary temperature were also compared.


A detailed methodology, specifically EIT analysis, sample size calculation and statistical approach are available in online supplementary material.

Study design

This prospective observational study was conducted in the Neonatal Intensive Care Unit (NICU) of the Royal Children’s Hospital Melbourne, a quaternary-level complex surgical and medical NICU.


Infants of any gestation were eligible if they were considered clinically suitable for SSC and were spontaneously breathing or receiving non-invasive respiratory support, continuous positive airway pressure or high-flow nasal cannula. Infants were not studied if receiving mechanical ventilation via an endotracheal tube or diagnosed with a disease that would affect the ability to interpret EIT data, such as chest wall abnormality.


Regional ventilation parameters were measured for 10 continuous minutes in each nursing period using the Pioneer EIT system sampling at 48 frames/s and a neonatal 32-electrode EIT belt (Sentec AG, Landquart, Switzerland).16 20–24 The EIT belt was placed on the infant at the start of the study and removed on completion of all measurements. HR, SpO2 and RR were manually recorded minutely from an Intellivue MP70 bedside monitor (Phillips Healthcare, Eindhoven, Netherlands). Axillary temperature was measured at the start and end of each measurement period.

Study intervention

Each infant acted as their own control. The study was divided into three measurement periods: pre-SSC (supine), during SSC and post-SSC (prone) (online supplementary figure 1). SSC consisted of nursing the nappy-clad infant prone on a parent chest while in a reclining chair at 30°, and infants were allowed to settle for 30 min in each position before measurements were taken.

Data analysis and outcomes

Data were analysed post hoc. EIT images were correctly orientated to the infant’s position (online supplementary figure 2).16 24 The first 30 clean, artefact-free, consecutive breaths from each 10 min recording were used for analysis. For each breath, the inspiratory time (Ti), mean centre of ventilation (CoV) in the right–left (RL) and ventrodorsal (VD) plane of the chest, percentage of total lung ventilation within the ventral, central and dorsal regions of the lung, and the percentage of lung with no apparent tidal ventilation were calculated.

Sample size and statistical analysis

A feasibility sample of 20 infants was chosen to detect a 2.3 effect size (3.6 SD, power 80% and alpha error 0.05). A p value of <0.05 was considered significant.


The characteristics of the 20 infants studied are summarised in table 1 and the Consort Diagram in the online supplementary figure 3. Eleven infants were preterm and nine infants were term. At the time of the study, infants presented with a range of clinical diagnoses representative of a quaternary NICU.

Table 1

Subject demographics and clinical data

Regional ventilation parameters

CoVVD and CoVRL demonstrated high intra-subject variability in all positions (online supplementary figures 4 and 5). Overall ventilation was greater in the right lung, with a mean CoVRL <43% (homogeneity) in all positions (figure 1A,B). There was no difference between CoVRL between any positions (p=0.10–0.63; mixed-effects model). CoVVD was also not consistent with homogeneous ventilation, favouring the dorsal lung in each position (figure 1C,D). Dorsal lung inhomogeneity was greater during SSC compared with prone; mean (95% CI) difference in CoVVD −2.0 (−0.4 to –3.6)%.

Figure 1

(A) CoVRL during supine (red symbols), skin-to-skin contact (SSC) (green symbols) and prone (blue symbols) positioning. Circles represent individual breath data from all 20 infants, and black bar represents mean±SD. Dashed line CoVRL representing purely homogeneous ventilation (43%). (B) Representative fEIT orientation image from infant 4. Magnitude of relative ventilation indicated using a hot-map (most ventilation in white and least in grey/dark blue). White dashed line illustrates the right–left plane, with the most-right lung regions at 0% and most-left at 100%. CoVRL associated with homogeneous ventilation (43%) represented by the red dot (C). CoVVD using the same symbols as panel A, with CoVVD of 55% representing homogeneous ventilation. (D) Representative fEIT orientation image from infant 4 using same descriptors as panel B. *p=0.013; mixed-effects model.

On further analysis by lung regions, there was less mean (95% CI) percentage of ventilation in the ventral region in both supine and prone compared with SSC, −2.0 (−0.3 to –3.6)% and −2.1 (−0.6 to –3.5)% respectively (table 2). In the dorsal region, ventilation in prone was −4.0 (−0.9, –7.2)% compared with SSC. SSC and prone positioning resulted in more lung regions without any apparent tidal ventilation compared with supine; mean (95% CI) difference between SSC and supine 5.0 (1.5 to 5.3)%, and 4.5 (3.9 to 5.1)% prone to supine (mixed-effects model). Online supplementary figure 6 shows the breath-by-breath data for each infant.

Table 2

Percentage of total lung ventilation in the ventral, central and dorsal regions of the lung, and percentage of apparently unventilated lung tissue

Other physiological parameters

There were no statistically or clinically significant differences in HR, RR or temperature between any of the positions (table 3). Ti was a mean (95% CI) 60 (10 to 160) ms longer during prone compared with supine; however, this was not deemed to be clinically relevant. There was a statistically significant but not clinically important decrease in SpO2 during SCC compared with both supine and prone; mean difference (95% CI) −1.3 (−0.6 to –2.0)% and −1.2 (−0.4 to –2.0)%, respectively. No infants experienced hypoxia (SpO2 <90%) during the measurement periods.

Table 3

Physiological measurement for all study periods


This is the first study to describe the complex regional ventilation patterns during SSC compared with cot-nursing. Irrespective of nursing position, there was a predominance of right and dorsal lung ventilation, with SSC resulting in greater dorsal lung ventilation compared with prone cot-nursing. The 2.0% difference in the CoVVD (geometric mean) represents a large shift in the distribution of VT throughout the chest,17 and has been previously associated with differences in lung injury.22 Nursing prone during SSC or in-cot increased the amount of unventilated lung compared with supine nursing. These patterns of regional ventilation occurred without differences in HR, RR and temperature, or adverse events.

Ventilation favouring the non–gravity-dependent lung (in supine this is the ventral lung and in prone the dorsal lung) is often reported in critical care settings and has been associated with worse respiratory status.20 25–27 This inhomogeneity is attributed to an intrathoracic pressure gradient resulting from the compression of dependent lung by the weight of the thoracic contents.25 Most reports of regional ventilation in critical care settings have been in patients with active acute respiratory disease and/or invasive mechanical ventilation.20 25–27 The infants receiving SSC in our study did not have active lung disease and dorsal dominance of ventilation was found irrespective of position. Dorsal-favoured ventilation as well as a limited influence of gravity have been reported in spontaneously breathing infants using EIT.18 19 Our observations support these findings and suggest that the pattern of ventilation in healthy infants is anatomically distributed rather than being gravity dependent as observed in infants with acute respiratory diseases.

Anatomical distribution does not explain the role of SSC in the greater dorsal inhomogeneity compared with prone positioning. There are several biological explanations for the greater dorsal ventilation during SSC compared with prone. First, during SSC, the infant is more upright; this may lessen the pressure of the abdominal contents on the diaphragm due to gravity and improve diaphragmatic contractility.14 As the diaphragm directly engages a greater portion of the dorsal lung, it may preferentially benefit dorsal ventilation.28 Second, during SSC, the infant is often nursed curved around the parental chest. This could lead to the alveoli in the ventral regions of the lung (closest to the parents chest) being more compressed than alveoli in the more dorsal parts of the lung (closest to the infant’s back). This difference in compression creates a ventral–dorsal thoracic pressure gradient not present during cot-nursing, with less pressure in the dorsal parts of the lung.

To our knowledge, this is the first report of the role of posture on the percentage of unventilated tissue, and the increase during SSC and prone was unexpected. The unventilated lung regions were in the distal lung. It is possible that the increased pressure of the abdomen on the thorax while prone increased the alveolar opening pressures, thus making it harder to maintain tidal volume in the distal lung regions supported by smaller airway branches.25 However, it is also possible that EIT reconstruction algorithm idiosyncrasies accounted for the findings.16

All positions were associated with stable HR, RR, SpO2, Ti and temperature, similar to other observational studies during SSC.9 11 13 Our study was not designed to assess harm or benefit; therefore, conclusions regarding the clinical impact of the observed regional ventilation patterns cannot be made. Regional ventilation patterns appeared to be affected during SSC, and this should be considered in future studies that examine harm/benefit associated with SSC. The association with general clinical stability reinforces the argument that in the stable, spontaneously breathing infant without lung disease, a degree of preferential dorsal ventilation may reflect normal physiology. Given these findings, this study supports the continuation of SSC in NICU environments.

Our study also provides insight into ventilation patterns that clinicians should be aware of. Specifically the potential difference in ventilation patterns between healthy infants and those with acute respiratory disease. Future research needs to consider the role of different infant populations and pathologies, as it is highly unlikely that there is a common ventilation pattern across all NICU infant populations and developmental states. This is fundamentally important for our understanding of respiratory care in the NICU. Traditional measures of respiratory function in the NICU are unable to identify the complex and dynamic behaviour of the infant lung. Importantly, detailed respiratory physiological measures could be made without impacting on normal clinical care or parent–infant bonding. EIT is the only tool with this ability available in the NICU.

There are limitations associated with this study. The standardised order of supine, SSC and then prone may have influenced results and randomising the order may negate any adaptation effects between positions. Analysis was not blinded to the position of the infant, which may have introduced bias; however, a standardised protocol for analysis, including pre-built EIT software algorithms, reduced this risk. The angle of the infant during SSC was difficult to control given the varying sizes of parental chests; this may have impacted both the cardiorespiratory parameters as well as the EIT model accuracy. Electrical impedance values are reconstructed using finite-element models of the thoracic contents based on libraries of CT images.16 Although this method of single-slice EIT is a validated measure of regional volumes,16 23 it is not without important limitations for this study. Infant chest models are based on the supine position. Prone position causes some chest wall deformity and may have influenced the boundaries of analysed lung regions. The position of the distal lung boundaries may account for the observed differences in the apparent unventilated lung tissue. The EIT child model of reconstruction used in this study also has a considerably smaller left lung, which may also explain the unexpected prominence of right lung ventilation.


SSC generates unique regional ventilation patterns compared with prone cot-nursing without changes to HR, SpO2, RR or temperature. Further investigation will be required to determine the impact, if any, of the greater dorsal lung ventilation observed during SSC. Gravity also plays a limited role on ventilation patterns with positional change in stable, healthy infants.


The authors acknowledge all of the infants and parents who agreed to participate in this study. We also acknowledge the nursing staff of the Butterfly Ward, RCH, for assistance with infant handling and care during the study.



  • Twitter @lmhickey, @DavidTingay

  • Contributors NFS, LH, DGT, EJP and PMP-F developed the concept, designed the experiment and interpreted the data. NFS, SK, IS, GD and EJP were involved in the data collection. NFS, LH and DGT supervised all aspects of the study and subsequent data analysis. NFS and GD performed the EIT analysis. All authors participated in data interpretation under supervision of NFS, LH and DGT. NFS wrote the first draft and all authors contributed to redrafting the manuscript.

  • Funding This study is supported by the Victorian Government Operational Infrastructure Support Programme (Melbourne, Australia). NFS, GD and SK were supported by a MCRI Honours Programme scholarship. DGT is supported by a National Health and Medical Research Council Clinical Career Development Fellowship (Grant ID 1053889) and the Royal Children’s Hospital Foundation. All EIT hardware was purchased by Murdoch Children’s Research Institute.

  • Competing interests None declared.

  • Patient consent for publication Not required.

  • Ethics approval Royal Children's Hospital Human Research Ethics Committee (HREC 36159B).

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Data availability statement All data, including raw data used for all figures and analysis, are available on reasonable request to the corresponding author.