Article Text

Download PDFPDF

Return of spontaneous circulation with a compression:ventilation ratio of 15:2 versus 3:1 in newborn pigs with cardiac arrest due to asphyxia
  1. Anne L Solevåg1,2,3,4,
  2. Ingrid Dannevig1,2,3,4,
  3. Myra Wyckoff5,
  4. Ola D Saugstad3,
  5. Britt Nakstad1,2
  1. 1Department of Paediatrics, Akershus University Hospital, Lørenskog, Norway
  2. 2Institute for Clinical Medicine, University of Oslo, Oslo, Norway
  3. 3Department of Paediatric Research, Oslo University Hospital, Rikshospitalet, Oslo, Norway
  4. 4Institute for Surgical Research, Oslo University Hospital, Rikshospitalet, Oslo, Norway
  5. 5Department of Pediatrics, University of Texas Southwestern Medical Center at Dallas, Dallas, Texas, USA
  1. Correspondence to Anne L Solevåg, Department of Paediatrics, Akershus University Hospital, 1478 Lørenskog and Department of Paediatric Research, Oslo University Hospital, Rikshospitalet, 0027 Oslo, Norway; a.l.solevag{at}


Objective International guidelines recommend a compression to ventilation (C:V) ratio of 3:1 in neonates, and 15:2 for other paediatric age groups. The authors aimed to compare these two C:V ratios in a neonatal swine model of cardiac arrest following asphyxia.

Design Experimental animal study.

Setting Facility for animal research.

Subjects 22 newborn pigs (age 12–36 h, weight 2.0–2.7 kg).

Interventions Progressive asphyxia until asystole. Animals were randomised to receive C:V 3:1 (n=11) or 15:2 (n=11).

Main outcome measures Return of spontaneous circulation (ROSC) was defined as a heart rate ≥100 bpm. Also of interest were haemodynamic parameters, cerebral and systemic oxygen saturation and the proinflammatory cytokine interleukin-1β (IL-1β).

Results Two animals in each group did not achieve ROSC. Mean (SD) increase in diastolic blood pressure (DBP; mm Hg) during compression cycles was significantly higher at a C:V ratio of 15:2 than 3:1 (7.1 (2.8) vs 4.8 (2.6)). Median time (IQR) to ROSC for the 3:1 group was 150 (140–180) s, and 195 (145–358) s for the 15:2 group. There were no significant differences in the temporal changes in haemodynamic parameters or oxygen saturation indices between the groups. IL-1β levels in cerebrospinal and bronchoalveolar lavage fluid was comparable between the groups.

Conclusion In neonatal pigs with asphyxia-induced cardiac arrest, the response to a C:V ratio of 15:2 is not better than the response to a C:V ratio of 3:1 despite better generation of DBP during resuscitation.

View Full Text

Statistics from


The 2010 guidelines of the International Liaison Committee on Resuscitation (ILCOR) acknowledge that the ideal ratio of chest compressions to ventilations (C:V ratio) during neonatal cardiopulmonary resuscitation (CPR) is unknown.1 Updated neonatal guidelines recommend that chest compressions should be given at a ratio of three compressions to one ventilation if the pulse is <60 bpm and not increasing despite at least 30 s of effective ventilation.2 Need for resuscitation in newborn infants is almost always the result of respiratory problems,3 and the reversal of asphyxia via adequate ventilation is critical in the case of asystole in the newborn. The cause of arrest in older age groups is more frequently ventricular fibrillation, and under these circumstances coronary perfusion pressure (CPP) is thought to be an important determinant of return of spontaneous circulation (ROSC) and neurological outcome.4,,7 Mathematical models and experimental data from adult animals indicate the beneficial effects of more cardiac compressions in a series,4 7,,10 and recommendations for older children and adults include longer compression sequences with less interruption for ventilation.11

What is already known on this topic

  • In adults and older children higher coronary perfusion pressures are generated with higher compression:ventilation ratios, leading to faster return of spontaneous circulation.

  • 9:3 Compressions to ventilations might not be better than 3:1, but no other scientific studies have been published on higher compression:ventilation ratios in neonates.

What this study adds

  • Fifteen compressions to two ventilations were not better than three compressions to one ventilation in this animal model of neonatal asphyxia and cardiac arrest.

  • The theory that newborns respond differently to resuscitation interventions than older children and adults is supported.

Our group has previously undertaken a study showing that a ratio of 9:3 was not better than 3:1 in a piglet model of neonatal asphyxia and cardiac arrest.12 However, whether longer sequences of uninterrupted compressions would result in better responses to neonatal CPR is unknown.

The aim of our current study was to compare an even longer compression cycle to standard neonatal CPR in the same piglet model of asphyxia-induced asystole. As 15:2 is the currently recommended ratio when CPR is performed by two or more healthcare personnel in all children except newborns, we chose to compare the effect of 15:2 versus 3:1 C:V on time to ROSC during neonatal CPR. Secondary outcomes of interest included the effect of compression cycle length on mean arterial blood pressure (MAP), diastolic blood pressure (DBP) as a rough estimate of CPP, acid/base status, the proinflammatory cytokine interleukin-1β (IL-1β), and systemic and regional cerebral oxygen saturation (rScO2). We hypothesised that a C:V ratio of 15:2 would lead to a faster ROSC than 3:1 during resuscitation from asphyxia-induced cardiac arrest. Secondary hypotheses were that use of the 15:2 ratio during neonatal CPR would be accompanied by better haemodynamic status and oxygen saturation indices following successful resuscitation with less evidence of inflammatory damage.


Experimental protocol

The experimental protocol (figure 1) was approved by the Norwegian Council for Animal Research. Animals were cared for and handled in accordance with the European guidelines for the use of experimental animals by certified category C researchers of the Federation of European Laboratory Animal Science Associations.

Figure 1

Experimental protocol. BALF, bronchoalveolar lavage fluid; CSF, cerebrospinal fluid; I.v., intravenous; MAP, mean arterial blood pressure.

Animal preparation

Twenty-two Noroc pigs of both sexes, 12–36 h of age (2.0–2.7 kg) were anaesthetised and mechanically ventilated (see appendix). As previously described,12 surgical placement of catheters for the measurement of arterial blood pressure, blood sampling and intravenous drug administration was carried out, and animals were stabilised on the ventilator for 1 h prior to induction of asphyxia.


Piglets were randomised to CPR using a C:V ratio of 3:1 (n=11) or 15:2 (n=11) and stratified according to gender.


Following baseline measurements, progressive asphyxia was induced by reducing oxygen in inspiratory gases to 8% and the ventilator rate by 10 breaths/min every 10 min. Also, as former experience with this animal model has shown that pCO2 does not reach asphyxial levels with hypoxia of this duration, CO2 was added in order to achieve a pCO2 of >7.0 kPa. Asphyxia was maintained until asystole occurred. Asystole was defined as a MAP of 0 mm Hg with loss of pulsatility and a flat ECG. Cardiac auscultation was also performed to confirm a heart rate of 0 bpm.

Resuscitation procedure

After 20 s of asystole, resuscitation was initiated by setting the ventilator to FiO2 0.21, peak inspiratory pressure 25 cm H2O and a rate 40 breaths/min for 30 s followed by coordinated cardiac compressions and ventilations at the ratio assigned by randomisation for 30 s. If asystole/severe bradycardia persisted, this was followed by intravenous adrenaline (0.02 mg/ kg). As per ILCOR guidelines, compressions and ventilations were continued and adrenaline administered every 3 min until a heart rate of ≥60 bpm was achieved.

The resuscitation protocol was implemented by a two-member resuscitation team. One person was assigned to each of the following roles: (1) chest compressions at a metronome-guided rate of 90 compressions/min. (chest compressions were aimed to generate a MAP of ≥20 mm Hg); and (2) administration of intravenous adrenaline and code supervisor who coordinated the timing and sequence of resuscitation interventions. ROSC was defined as an unassisted pulse rate of ≥100 bpm. If there was no ROSC after 15 min of resuscitation, the procedure was ended. If ROSC was achieved, the animal was maintained on the ventilator for 4 h prior to euthanasia. Following ROSC, if animals were developing severe hypotension (MAP <15 mm Hg), they were given a bolus of normal saline 10 ml/kg. In the case of seizures, a bolus of midazolam (5 mg/ml) 2.5 mg/kg was administered.

Four hours after ROSC, just prior to euthanasia, cerebrospinal fluid (CSF) was collected through a lumbar puncture, and the animal was euthanised. Bronchoalveolar lavage was then performed using 30 ml/kg warm saline (38°C). Aspirated samples were centrifuged at 2000g and 4°C for 20 min to remove cells. The supernatant was transferred to polypropylene tubes and frozen at −80°C for cytokine analysis.


Heart rate, arterial blood pressure, ECG and SpO2 were continuously monitored and recorded using Biopac modules (model MP 150; Biopac Systems, Goleta, California, USA). Arterial blood samples (0.2 ml) for temperature corrected blood gas, glucose and lactate analysis (Blood Gas Analyser 860; Ciba Corning Diagnostics, Medfield, Massachusetts, USA) were drawn from the catheter in the common carotid artery immediately following surgical instrumentation (ie, at the beginning of the 60 min stabilisation period) and after 20, 40 and 60 min (ie, at the end of the stabilisation period, defined as ‘baseline’). Arterial blood gases were also analysed every 5 min throughout asphyxiation, and after each of the first 5 min following ROSC. Thereafter, the intervals were gradually prolonged, and during the last 3 h of post-resuscitation observation, arterial blood gases were analysed half-hourly.

Rectal temperature was continuously monitored with a flexible digital thermometer (Ama-digit ad 15 th; Amarell Electronic, Kreuzwertheim, Germany) and maintained between 38°C and 40°C with a heating blanket and a radiant heat lamp. End-tidal CO2 was monitored using a Normocap Oxy capnometer (Datex, Helsinki, Finland).

Near infrared spectroscopy

A probe (Pediatric SomaSensor; Somanetics, Troy, Michigan, USA) was placed on the left side of the skull, contralateral to the catheter in the common carotid artery. rScO2 was continuously monitored and recorded. rScO2 is a relative parameter which cannot be used for comparison between individuals. We therefore calculated the changes in rScO2 expressed as the rScO2 actual/rScO2 baseline. Cerebral fractional tissue oxygen extraction (cFTOE) was calculated from rScO2 and SpO2 values. The ratio of (SpO2−rScO2)/SpO2 represents the balance between oxygen delivery and oxygen consumption.13

Cytokine analysis

IL-1β in bronchoalveolar lavage fluid (BALF) and CSF was measured to assess inflammation resulting from hypoxia and reoxygenation/reperfusion. The cytokine was analysed in a blinded fashion using the Quantikine Porcine IL-1β Immunoassay (R&D Systems Europe, Abingdon, UK). All samples were analysed in duplicate following the manufacturer's instructions for cell culture supernatant samples. Samples were measured spectrophotometrically at 450 nm with wavelength correction set to 540 nm (Varioskan; Thermo Fisher Scientific, Waltham, Massachusetts, USA).

Sample size and data analysis

To determine the number of pigs needed, a power analysis was performed based on data from a previous study comparing room air with 100% oxygen in the same piglet model.14 In order to detect a difference of 50 s (determined as a ‘clinically significant’ difference after a pilot study) in time to ROSC between treatment arms with a power of 80% and a type I error rate of 5%, 11 pigs were needed for each treatment group.

Statistical analysis was performed using SPSS 15.0 for Windows. Descriptive statistics are reported as mean and SD or median and IQR. The Student t test was used for continuous variables for comparisons between groups, except for times to ROSC where the Mann–Whitney test was used. Fisher's exact test was used to compare categorical variables.

To investigate the temporal changes in a set of variables such as SpO2, we fitted repeated measurement models. These are analysis of variance (ANOVA) models with two factors, time and intervention (group), accommodating dependent time observations for each pig, but independence between the pigs. The results from the repeated measurement models are tests for differences between the groups, points of time, and interaction between time and group. p Values <0.05 were considered statistically significant.


Two of the animals in each group resuscitated with a C:V ratio of 3:1 and 15:2, respectively, did not achieve ROSC and were excluded from analysis.

Baseline characteristics

Animal characteristics at baseline are shown in table 1. No significant differences were noted regarding age, sex distribution, heart rate, MAP, arterial blood gases, glucose, lactate, SpO2 or haemoglobin between the groups.

Table 1

Characteristics at baseline


Table 2 compares animal characteristics at the time of asystole. The duration of hypoxia/asphyxia needed to achieve asystole was comparable between the groups as were arterial blood gases, glucose and lactate.

Table 2

Characteristics at asystole


The number of cardiac compressions delivered per minute was significantly higher at a C:V ratio of 15:2 compared to 3:1 (table 3). Also, mean increase in DBP (mm Hg) during compression cycles was significantly higher for 15:2 than 3:1 (7.1 (2.8) vs 4.8 (2.6), p=0.004).

Table 3

Characteristics of resuscitation and immediately following return of spontaneous circulation (ROSC)

Return of spontaneous circulation

Median time (IQR) to ROSC for the 3:1 group was 150 (140–180) s, and 195 (145–358) s for the 15:2 group, with no significant difference between the groups. Haemodynamic parameters and arterial blood gases immediately after ROSC were similar for the two groups (table 3).


Piglets resuscitated with a C:V ratio of 3:1 and 15:2 had no significant differences in mean IL-1β in BALF or CSF (table 4).

Table 4

Characteristics at euthanasia (4 h after resuscitation)

Blood gases

There were no significant differences in temporal changes in pH, pCO2 or base excess between the groups.

Oxygen saturations and consumption

Temporal changes in SpO2 throughout the experiment were similar between the groups. There were no significant differences between the groups in temporal changes in rScO2 actual/rScO2 baseline or cFTOE.

Haemodynamic parameters

Temporal changes in heart rate and MAP throughout the experiment were similar between the groups.


In this study, there were no significant differences in time to ROSC, haemodynamic parameters after successful resuscitation, oxygen saturation indices, or IL-1β in CSF and BALF when providing CPR at a C:V ratio of 15:2 versus 3:1.

In adult animal models of cardiovascular collapse due to ventricular fibrillation where different C:V ratios were compared, invasive pressure measurements and calculations have been conflicting. Despite the theoretical advantage regarding pressures when more compressions in a series are delivered, Dorph et al found no difference in CPP when pigs were resuscitated at a C:V ratio of 30:2 or with compressions only.15 On the ether hand, Kern et al measured both a higher aortic diastolic pressure and CPP at 4 min of resuscitation with compression-only CPR compared to a C:V ratio of 15:2,8 and Ewy and co-workers showed that integrated CPP was higher with uninterrupted chest compressions than CPR at a C:V ratio of 30:2.16 In our study DBP was determined as an indirect measure of CPP during the resuscitation sequence. We demonstrated a significantly higher DBP resulting from 15 as compared to three compressions in a series; however, neither method reached the threshold of 15–20 mm Hg thought to be critical to achieve ROSC.17 One possible explanation for the failure to achieve adequate DBP might be massive systemic vasodilatation resulting from the severe acidaemia in this model. Another possible explanation is ineffective chest compressions. Chest compressions have been shown to be less effective when the rate is too low, the chest is stiff and/or the chest is large or oddly shaped (eg, barrel chests in adults with chronic lung disease). As we will discuss, the rate of compressions in this model was too low at either ratio, and this might have affected the poor increase in DBP.

Our animal model mimics acute asphyxia in the term newborn baby where time from start of the insult to terminal apnoea is thought to be approximately 20 min,18 comparable to the mean duration of hypoxia in this study (table 2). However, a weakness of our study was that the animals had undergone transition from intra- to extrauterine life. Another weakness is that four animals were excluded from analysis because of failure to attain ROSC. This left us with only nine animals in each group, raising the possibility of a type II error due to too small sample sizes. However, when we performed the statistical experiment of increasing the 3:1 group by including ROSC data from our previous study comparing the C:V ratios of 3:1 and 9:312 in the very same piglet model, median time to ROSC was not significantly different between 3:1 (n=24) and 15:2 (n=9).

The number of compressions delivered per minute was significantly higher at a ratio of 15:2 than 3:1, and fairly low in the 3:1 group (table 3). This finding is supported by a manikin study that showed that the current infant and child C:V ratios (3:1 and 5:1 at the time of this study) resulted in fewer chest compressions and more rescue ventilations than the higher adult (15:2) ratio.19 Interestingly, neither group in our study achieved the recommended 90 compressions per minute (table 3), which suggests that time spent performing adequate ventilations was interfering with compressions at a C:V ratio of 15:2, as well as 3:1.

A mathematical analysis of optimal C:V ratios in non-neonatal paediatric and adult CPR based on theoretical oxygen delivery and blood flow, concluded that current guidelines overestimate the need for ventilation during standard CPR by two- to fourfold.20 In this paper, Babbs and Kern point out that for asphyxial arrest one needs to attempt several ventilations initially to restore alveolar gas concentrations towards normal before performing compressions and ventilation at the calculated C:V ratio. In our study this was done by providing an initial 30 s of positive pressure ventilation before initiation of cardiac compressions. The hypoxaemia that usually precedes cardiac arrest in the paediatric population is taken into consideration in another mathematical analysis by Babbs where an optimal C:V ratio of 5:1 was found for trained personnel resuscitating a child weighing 5 kg,10 which might be an approximation to a newborn.

Despite the conclusions of Babbs and Nadkarni that the C:V ratio of 15:2 is too high in the lowest age groups,10 we wanted to investigate this ratio in our newborn animal model of cardiac arrest following asphyxia. As it has been recognised that the mastery and retention of CPR skills is improved by simplifying the technique and algorithms taught,21 22 recommendations for a ratio of 15:2 for all paediatric age groups would make paediatric CPR easier to teach, learn, remember and perform. In addition, excessive ventilation might impair cardiac output and decrease coronary and cerebral perfusion pressures in adults.23 24 However, since the aetiology of cardiovascular collapse is invariably asphyxia in the newborn rather than the ventricular fibrillation of the adult, one must be cautious in decreasing ventilation too much. Also, due to the unique physiology of the newborn where proper aeration of the lungs is required for promoting postnatal adaptation,25 data from adult studies might not be transferrable to the neonate.

The study design and results only allow us to suggest that 15:2 is not better than 3:1 in neonatal resuscitation. Whether an extended series of cardiac compressions is as good as 3:1 cannot be concluded from this study. However, the intended number of cardiac compressions is particularly difficult to achieve at a C:V ratio of 3:1. Further studies should examine if alternative C:V ratios offer advantages over the currently recommended 3:1.

In conclusion, a C:V ratio of 15:2 did not give a faster ROSC than 3:1 in newborn pigs with cardiac arrest resulting from asphyxia. Also, there were no significant differences in haemodynamic parameters after successful resuscitation, oxygen saturation indices, or IL-1β in CSF and BALF between the groups. The lack of an evidence base for current recommendations implies that further studies investigating different C:V ratios in neonates should be undertaken.


The authors thank Camilla Skjæret and Tonje Sonerud for help in optimising cytokine measurements and Geir Aamodt for statistical assistance. The authors express gratitude to Respiratory Covidien Norway AS for the loan of an INVOS cerebral oximeter.


View Abstract


  • Funding The authors thank the Laerdal Foundation for Acute Medicine and the University of Oslo's Foundation at Akershus University Hospital for significant financial contributions to the study.

  • Competing interests None.

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

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.