Heart-rate complexity for prediction of prehospital lifesaving interventions in trauma patients

Abstract

Background: Traditional vital signs often fail to identify critically injured patients soon enough to permit timely intervention. To improve our ability to forecast the need for prehospital lifesaving interventions (LSIs), we applied heart-rate complexity (HRC) analysis to the electrocardiogram (ECG) of patients en route to trauma centers.

Methods: Analysis of ECG and clinical data from 374 patients en route by helicopter to three urban Level I trauma centers was conducted. Waveforms from 182 patients were excluded (because of ectopy, noise, or inadequate length). Of the remaining 192 patients, 54 received 66 LSIs in the field (LSI group): intubation (n = 52), cardiopulmonary resuscitation (n = 5), cricothyroidotomy (n = 2), and pneumothorax decompression (n = 7); 138 patients did not (non-LSI group). In the field, heart rate, blood pressure, and the Glasgow Coma Scale score (GCS(TOTAL)) and its motor component (GCS(MOTOR)) were recorded. ECG was recorded during flight. Ectopy-free, 800-beat sections of ECG were identified off-line and analyzed by HRC methods including Sample Entropy (SampEn) and Detrended Fluctuations Analysis (DFA).

Results: There was no difference between LSI and non-LSI patients in heart rate or blood pressure. SampEn was lower in LSI than in non-LSI (0.88 +/- 0.03 vs. 1.11 +/- 0.03), as was DFA (1.09 +/- 0.05 vs. 1.33 +/- 0.03) and GCS(MOTOR) (3.4 +/- 0.4 vs. 5.7 +/- 0.1) (all p < 0.0001). By logistic regression, SampEn, DFA, and GCS(MOTOR) were independently associated with LSIs (area under the receiver operating characteristic curve, 0.897).

Conclusions: Decreased HRC is associated with LSIs in prehospital trauma patients. HRC may be useful as a new vital sign for identification of the severely injured.