Delivering manual cardiopulmonary resuscitation (CPR) in a diving bell: an analysis of head-to-chest and knee-to-chest compression techniques.

Introduction: Chest compression often cannot be administered using conventional techniques in a diving bell. Multiple alternative techniques are taught, including head-to-chest and both prone and seated knee-to-chest compressions, but there are no supporting efficacy data. This study evaluated the efficacy, safety and sustainability of these techniques. Methods: Chest compressions were delivered by a team of expert cardiopulmonary resuscitation (CPR) providers. The primary outcome was proportion of chest compressions delivered to target depth compared to conventional CPR. Techniques found to be safe and potentially effective by the study team were further trialled by 20 emergency department staff members. Results: Expert providers delivered a median of 98% (interquartile range [IQR] 1.5%) of chest compressions to the target depth using conventional CPR. Only 32% (IQR 60.8%) of head-to-chest compressions were delivered to depth; evaluation of the technique was abandoned due to adverse effects. No study team member could register sustained compression outputs using prone knee-to-chest compressions. Seated knee-to-chest were delivered to depth 12% (IQR 49%) of the time; some compression providers delivered > 90% of compressions to depth. Conclusions: Head-to-chest compressions have limited efficacy and cause harm to providers; they should not be taught or used. Prone knee-to-chest compressions are ineffective. Seated knee-to-chest compressions have poor overall efficacy but some providers deliver them well. Further research is required to establish whether this technique is feasible, effective and sustainable in a diving bell setting, and whether it can be taught and improved with practise.