Dataset Information

Automated left atrial strain analysis for predicting atrial fibrillation in severe COVID-19 pneumonia: a prospective study.

ABSTRACT:

Background

Atrial fibrillation (AF) is the most documented arrhythmia in COVID-19 pneumonia. Left atrial (LA) strain (LAS) analysis, a marker of LA contractility, have been associated with the development of AF in several clinical situations. We aimed to assess the diagnostic ability of LA strain parameters to predict AF in patients with severe hypoxemic COVID-19 pneumonia. We conducted a prospective single center study in Amiens University Hospital intensive care unit (ICU) (France). Adult patients with severe or critical COVID-19 pneumonia according to the World Health Organization definition and in sinus rhythm were included. Transthoracic echocardiography was performed within 48 h of ICU admission. LA strain analysis was performed by an automated software. The following LA strain parameters were recorded: LA strain during reservoir phase (LASr), LA strain during conduit phase (LAScd) and LA strain during contraction phase (LASct). The primary endpoint was the occurrence of AF during ICU stay.

Results

From March 2020 to February of 2021, 79 patients were included. Sixteen patients (20%) developed AF in ICU. Patients of the AF group were significantly older with a higher SAPS II score than those without AF. LAScd and LASr were significantly more impaired in the AF group compared to the other group (- 8.1 [- 6.3; - 10.9] vs. - 17.2 [- 5.0; - 10.2] %; P < 0.001 and 20.2 [12.3;27.3] % vs. 30.5 [23.8;36.2] %; P = 0.002, respectively), while LASct did not significantly differ between groups (p = 0.31). In a multivariate model, LAScd and SOFA cv were significantly associated with the occurrence of AF. A LAScd cutoff value of - 11% had a sensitivity of 76% and a specificity of 75% to identify patients with AF. The 30-day cumulative risk of AF was 42 ± 9% with LAScd > - 11% and 8 ± 4% with LAScd ≤ - 11% (log rank test P value < 0.0001).

Conclusion

For patients with severe COVID-19 pneumonia, development of AF during ICU stay is common (20%). LAS parameters seem useful in predicting AF within the first 48 h of ICU admission.

Trial registration

NCT04354558.

SUBMITTER: Beyls C

PROVIDER: S-EPMC8649321 | biostudies-literature |

REPOSITORIES: biostudies-literature

ACCESS DATA

Similar Datasets

Project description:BackgroundCurrent ablation therapy for atrial fibrillation is suboptimal, and long-term response is challenging to predict. Clinical trials identify bedside properties that provide only modest prediction of long-term response in populations, while patient-specific models in small cohorts primarily explain acute response to ablation. We aimed to predict long-term atrial fibrillation recurrence after ablation in large cohorts, by using machine learning to complement biophysical simulations by encoding more interindividual variability.MethodsPatient-specific models were constructed for 100 atrial fibrillation patients (43 paroxysmal, 41 persistent, and 16 long-standing persistent), undergoing first ablation. Patients were followed for 1 year using ambulatory ECG monitoring. Each patient-specific biophysical model combined differing fibrosis patterns, fiber orientation maps, electrical properties, and ablation patterns to capture uncertainty in atrial properties and to test the ability of the tissue to sustain fibrillation. These simulation stress tests of different model variants were postprocessed to calculate atrial fibrillation simulation metrics. Machine learning classifiers were trained to predict atrial fibrillation recurrence using features from the patient history, imaging, and atrial fibrillation simulation metrics.ResultsWe performed 1100 atrial fibrillation ablation simulations across 100 patient-specific models. Models based on simulation stress tests alone showed a maximum accuracy of 0.63 for predicting long-term fibrillation recurrence. Classifiers trained to history, imaging, and simulation stress tests (average 10-fold cross-validation area under the curve, 0.85±0.09; recall, 0.80±0.13; precision, 0.74±0.13) outperformed those trained to history and imaging (area under the curve, 0.66±0.17) or history alone (area under the curve, 0.61±0.14).ConclusionA novel computational pipeline accurately predicted long-term atrial fibrillation recurrence in individual patients by combining outcome data with patient-specific acute simulation response. This technique could help to personalize selection for atrial fibrillation ablation.