Chest compression-associated injuries in cardiac arrest patients treated with manual chest compressions versus automated chest compression devices (LUCAS II) - a forensic autopsy-based comparison

摘要

The aim of this autopsy study was to investigate chest-compression associated injuries to the trunk in out-of-hospital and in-hospital non-traumatic cardiac arrest patients treated with automated external chest compression devices (ACCD; all with LUCAS II devices) versus exclusive manual chest compressions (mCC). In this retrospective single-center study, all forensic autopsies between 2011 and 2017 were included. Injuries following cardiopulmonary resuscitation (CPR) in patients treated with mCC or ACCD were investigated and statistically compared using a bivariate logistic regression. In the seven-year period with 4433 autopsies, 614 were analyzed following CPR (mCC vs. ACCD: n = 501 vs. n = 113). The presence of any type of trunk injury was correlated with longer resuscitation intervals (30 +/- 15 vs. 44 +/- 25 min, p 0.05). In comparison with mCC, treatment with ACCD led to more frequent skin emphysema (5 vs 0%, p = 0.012), pneumothorax (6 vs. 1%, p = 0.008), lung lesions (19 vs. 4%, p = 0.008), hemopericardium (3 vs 1%, p = 0.025) and liver lesions (10 vs. 1%, p = 0.001), all irrespective of confounding aspects. Higher age and longer CPR durations statistically influenced frequency of sternal and rib fractures (p 0.001). The mean number of fractured ribs did not vary significantly between the groups (6 +/- 3 vs. 7 +/- 2, p = 0.09). In this cohort with unsuccessful CPR, chest compression-related injuries were more frequent following ACCD application than in the mCC group, but with only minutely increased odds ratios. The severity of injuries did not differ between the groups, and no iatrogenic injury was declared by the forensic pathologist as being fatal. In the clinical routine after successful return of spontaneous circulation a computed tomography scan for CPR-associated injuries is recommended as soon as possible.