Background. Pulmonary Embolism (PE) produces electrocardiographic (ECG) abnormalities in proportion to the degree of pulmonary hypertension and pulmonary vascular obstruction. We hypothesized that an ECG score may be of use, in combination with clinical parameters and prediction rules, in identifying such patients at risk for in-hospital clinical adverse events.
Methods. All adult patients admitted for at least 72 hours at the private and service wards of Philippine Heart Center (PHC) between January 2005 and November 2009 with confirmed PE were enrolled in the study. A clinical decision scoring system based on the Revised Geneva Score comprising of eight variables were used to stratify patients into high clinical probability or non-high clinical probability (Iow and intermediate) of PE. Electrocardiograms were obtained within 72 hours after suffering symptoms related to PE or onset of clinical deterioration after initial stabilization during admission. The ECG score was then calculated. Eligible patients were further subdivided into groups comprising of those with ECG score of ~ 3 and those with score of ~ 3. The information on the clinical outcomes were analyzed and compared among the four groups.
Results. One hundred (100) patients were included in the study. Twenty-one patients (21%) died during the index confinement. The incidence of hemodynamic deterioration and major complications were also found to occur in 37% and 20% of patients, respectively. The incidence of hemodynamic deterioration exhibited significant differences between groups with low and high ECG scores (19% vs. 47%, p=0.005). The incidence of in-hospital hemodynamic collapse and the necessity for treatment upgrade and catecholamine infusion were all significantly more frequent in patients with high ECG scores (11% vs. 30%, p=0.027; 17% vs. 38%, p=0.023; and 11% vs. 28%, p=0.039, respectively). The composite outcome of major complications did not reach statistical significance but recurrent thromboembolic events were noted to be significantly higher in patients with high ECGscore (3% vs. 17%, p=0.029). Mortality rate was also increased among patients with high ECG score (11% vs. 27%, p=0.050). In the subgroup analysis, patients with high clinical probability and high ECG scores in combination have a significantly higher incidence of hemodynamic deterioration compared to other groups (p=O.018). There was also a similar trend for this group to have an increased rate of mortality and major complications, although the difference did not reach statistical significance in this study.
Conclusions. The ECG-scoring system can be very useful in identifying patients at risk for developing clinical end points of mortality and hemodynamic deterioration among patients with low and high clinical probability of PE. Although the prognostic accuracy of the ECG score does not allow identification of all those who will develop adverse outcomes, it can provide an incremental role to the clinical stratification provided by the Revised Geneva Scoring.