Results
Typical carotid blood flow and arterial pO2 dynamics for OAC-CPR and CC-CPR are shown in figure 4. The flow waveform for OAC-CPR does not exhibit the rapid spikes that are evident with CC-CPR. A decrease in pO2 is evident over the course of the fibrillation episode for both types of CPR. Larger oscillations in the arterial pO2 at the start of each run are due to greater sensitivity to environmental conditions at higher O2 levels.
(Enlarge Image)
Figure 4.
Examples of (a) the flow waveform for the right carotid artery flow rate and (b) the change in arterial oxygen partial pressure as read by the FOXY sensor.
Using the GLM, differences in treatment effects were determined as follows. The intersubject variation significantly affected all metrics except carotid flow; run sequence was significant for EtCO2 and cardiac output. Rebreathing significantly increased EtCO2 by 36.6 mm Hg, and non-significantly increased time to ROSC by 39 s (SE=25.3 s). The interaction effect of run sequence and rebreathing was significant for the metric of EtCO2 and the time until ROSC, while the interaction effect of protocol and run sequence was significant for the metric of MAV. The first interaction effect likely points to a change in the lung architecture over time, affecting the effect of the rebreathing gas, most likely due to pulmonary oedema/broken ribs. The second interaction effect likely points to the fact that spontaneous respirations were repressed in the second study.
The major findings of the GLM for the CPP, carotid blood flow, MAV and cardiac output are indicated graphically in figure 5. CPP and MAV were greater for OAC-CPR than CC-CPR, although no difference was found for cardiac output. Carotid flow rate, however, was higher for CC-CPR with an average at 50 ml/min compared with 40.7 ml/min for OAC-CPR.
(Enlarge Image)
Figure 5.
Comparisons of datasets used in the general linear model for mechanical metrics. Error bars denote SD, and the sample size in each set is given near the bottom of the bars. Asterisks denote a significant difference between cardiopulmonary resuscitation (CPR) treatment effects (p<0.05), with the difference between the only abdominal compression (OAC)-CPR and the chest compression (CC)-CPR treatment effects (OAC effect minus CC effect) given in parentheses.
The major findings of the GLM for gas exchange and resuscitation metrics are shown in figure 6. End-tidal CO2 was significantly lower during OAC-CPR than CC-CPR and was comparable with the level during hyperventilation for protocol 2. Arterial pO2 decreased by nearly 35% during CC-CPR but only 12% for OAC-CPR. However, it was qualitatively more difficult to resuscitate subjects after OAC-CPR, as quantified by a greater delay in ROSC (+70 s).
(Enlarge Image)
Figure 6.
Comparisons of datasets used in the general linear model for metrics of gas exchange and ease of resuscitation. Error bars denote SD, and the sample size in each set is given near the bottom of the bars. Asterisks denote a significant difference between cardiopulmonary resuscitation (CPR) treatment effects (p<0.05), with the difference between the only abdominal compression (OAC)-CPR and the chest compression (CC)-CPR treatment effects (OAC effect minus CC effect) given in parentheses.
For some subjects, unexpected negative CPP values were observed during CC-CPR. We hypothesized that such values were due to permanent physical changes in the chest structure. To investigate, the aortic and right atrial pressures in the first CC-CPR run for each subject were analysed, and divided by whether or not a negative CPP was observed during that run. The divided sets were compared by t test, as shown in Table 1. The mean pressures for the first OAC-CPR run were also calculated. While the mean aortic pressure did not differ between the groups, the mean right atrial pressure was significantly higher for subjects in which a negative CPP was observed. Additionally, 4 out of 5 of the subjects with a negative initial CPP experienced broken ribs, compared with 5 of the 11 subjects that had a positive initial CPP.
A necropsy was performed on each of the 17 subjects after the completion of all runs. Of these 17, only one showed signs of abdominal damage. It should be noted that the abdominal damage (liver contusions and a ruptured bladder) was observed after 20 min of OAC-CPR and 40 min of CC-CPR were applied attempting to resuscitate the subject; 12 out of 17 of the subjects had significant remodelling or broken ribs, and 16 of the 17 had pulmonary oedema. Some remodelling of the chest was normally observed following the first run of CC-CPR in the subjects with such damage.