TEE-Guided CPR in Out-Of-Hospital Cardiac Arrest

Key Takeaways

• Sustained ROSC was similar across groups despite image-guided compression-site adjustment.
• Higher intra-CPR end-tidal carbon dioxide was observed during minutes 11 to 20 after emergency department arrival in the TEE-guided group.
• TEE imaging and compression-site relocation were achieved in many patients, and adverse event rates were comparable between groups.

In the EXECT-CPR randomized clinical trial, adults transported to the emergency department with ongoing resuscitation for nontraumatic out-of-hospital cardiac arrest had sustained ROSC in 44% with TEE-guided CPR and 39% with conventional CPR. Investigators tested whether transesophageal echocardiography could redirect compressions away from aortic valve compression and toward the left ventricle during resuscitation. The between-group difference in the primary outcome was not significant.

This single-center cluster-randomized clinical trial with periodic crossover was conducted in the emergency department of a tertiary medical center in Taipei, Taiwan, from June 26 to November 19, 2023. It enrolled 132 adults aged 20 years or older who presented to the emergency department with ongoing resuscitation for nontraumatic out-of-hospital cardiac arrest, with 66 patients assigned to each group. Median age was 68 years, and 66% were male. After intubation, TEE-guided CPR was used to identify aortic valve compression and redirect compressions toward the left ventricle, whereas controls received guideline-directed compressions at the lower half of the sternum. The prespecified primary outcome was sustained ROSC lasting at least 20 minutes.

The cluster-adjusted odds ratio for sustained ROSC was 1.21 with a 95% CI of 0.64 to 2.29. Any ROSC occurred in 33 of 66 patients in both groups, and survival to ICU admission was 20 of 66 in both groups. During the 11- to 20-minute interval after ED arrival, mean intra-CPR end-tidal carbon dioxide was higher with TEE guidance by 8.61 mm Hg. The proportion reaching etCO₂ above 20 mm Hg at any point was not statistically different, with a cluster-adjusted odds ratio of 1.54 (95% CI, 0.45 to 5.25), although etCO₂ data were unavailable in 45% of patients. Clinical outcomes remained similar between groups.

Before ROSC, TEE images were obtained in 52 patients in the intervention group, with a median 5.3 minutes from ED arrival to first image. The area of maximal compression was identified in 45 patients, and 36 required compression-site adjustment. Successful relocation occurred in 33 of those 36 patients, with completion at a median 8.5 minutes from ED arrival and 3 repositionings per adjustment. Operator interpretation showed strong agreement with expert review, with kappa 0.888, and mean end-tidal carbon dioxide rose from 23.5 to 28.3 mm Hg during successful adjustment. These procedures were often completed during ongoing resuscitation.

Investigators also reported that the trial was underpowered because anticipated effect-size assumptions were optimistic. TEE-related and CPR-related adverse event rates were comparable, and chest compression fraction was similar between groups. In the TEE-guided arm, one trivial pneumothorax required no intervention and one upper gastrointestinal bleed required treatment.