INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Electrocardiographic findings that relate to the severity of obstructive airway disease have been described in adults (1). Several investigators have studied correlates of spirometry with a standard 12-lead electrocardiogram (ECG). In studies correlating ECGs with spirometry, the best discriminant for airway obstruction was the mean frontal vector of the P wave (3). There is "verticalization" of the P axis with progressive hyperinflation. This is because the orientation of the atria changes with progressive hyperinflation as their inferior surface rests on the diaphragm (10). When the diaphragm descends, rotational forces on the atria cause changes in the P axis.

There have been no reported studies examining the relationship between the P axis and lung volumes in children. This is important because age-related changes in thoracic cavity shape and mechanical properties play an important role in airway dynamics (11). Thus hyperinflation may be seen early and more frequently in children with obstructive lung disease due to a more compliant chest wall. We hypothesized that a vertical P axis vector due to hyperinflation would be present in children even with mild airway obstruction and might serve as a reliable marker of air-trapping. Furthermore, we hypothesized that the vertical orientation of the P axis would be reversed with a bronchodilator, as airway obstruction and hyperinflation were relieved. Therefore we studied a group of children with obstructive airway disease before and after bronchodilator.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

All children presenting to the pediatric pulmonary function laboratory with the clinical diagnosis of asthma between October 1995 and May 1996 were prospectively studied. The local institutional review board approved the protocol for the study. Inclusion criteria included: age  6 years; ability to perform reproducible pulmonary function tests, i.e., coefficient of variation < 5% between maneuvers (12); reversible obstruction (defined as FEV₁ and FEF_25-75% of  80% predicted for age after bronchodilator). Exclusion criteria included subjects with congenital or acquired heart disease, arrhythmias, restrictive lung disease, or electrocardiographic/laboratory evidence of cor pulmonale. All subjects were asked not to use any bronchodilators for at least 12 h prior to the study.

After obtaining informed consent, a standard 12-lead ECG was performed in the sitting position. Subjects then underwent pulmonary function testing including spirometry (P. K. Morgan 100-L dry rolling seal spirometer) as per ATS guidelines (12) and estimation of lung volumes by whole body plethysmography using a whole-body pressure plethysmograph (P. K. Morgan) using standard techniques (15- 17). All pulmonary function tests were performed by the same technician (RG). The coefficient of variation for measurements of thoracic gas volume (TGV) for an individual patient in our laboratory is 5.5%.

The ECG and pulmonary function measurements were repeated at least 15 min after aerosolized inhalation of 2.5 mg of albuterol sulfate with 3 ml normal saline administered using a small volume hand-held nebulizer (Pulmoaide; Devilbiss, Somerset, PA). Lung function measurements were expressed as a percentage of predicted for age, height, and gender (12). Hyperinflation and bronchodilator response was assessed based on recommendations by the Intermountain Thoracic Society/Pulmonary Function Standardization Task Force and the American Thoracic Society (15, 16).

All ECGs were interpreted by a blinded investigator (J.S.). The mean frontal P axis was calculated to the nearest five degrees using Einthoven's method (18). P axis  60° was considered "vertical," axes < 40° and those between 40°-59° were considered horizontal and intermediate, respectively. Fifty ECGs were reanalyzed by the same investigator (J.S.) 1 mo later, and the intraobserver variability was < 5%.

Statistical Analysis

Data were analyzed using Microsoft Excel Version 7.0 Analysis Toolpak (Grey Matter International, Cambridge, MA). All means are expressed as ± SEM. The means for FVC, FEV₁, FEF_25-75%, and TGV are expressed as percentage predicted for age. The results before and after bronchodilator were compared using Student's paired t tests. Sensitivity, specificity and predictive value were calculated for P axis  60° as a measure of significant hyperinflation. Sensitivity was defined as the proportion of subjects with significant hyperinflation (defined as TGV  130% of predicted value) (16) who had a P axis  60°; specificity was defined as the proportion of subjects without significant hyperinflation who did not have a P axis  60°. The positive predictive value was defined as the proportion of subjects with P axis  60° who had significant hyperinflation, and the negative predictive value defined as the proportion of subjects without a P axis  60° who did not have significant hyperinflation. Sensitivity, specificity, and predictive values were also calculated for a decrease in P axis after bronchodilator as a measure of a decrease in hyperinflation. Response to bronchodilator and corresponding ECG changes in P and QRS axes were compared using the chi-square test. A p value of  0.05 was considered statistically significant.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

One hundred and two children met the study criteria. The mean age of the subjects was 10.5 ± 0.3 yr (range, 6 to 18 yr), with male:female ratio of 1.24:1. The spirometry, lung volumes, and ECG data are summarized in Table 1.

There was a wide range of lung function. Twenty-seven subjects (26%) had moderate to severe hyperinflation (TGV > 130% of predicted value) before bronchodilator. The P axis ranged from 10° to 80° with a mean of 54.9 ± 1.5°, and the QRS axis ranged from 0° to 120° with a mean of 59.3° ± 2.4°. After bronchodilator, the mean TGV decreased to 96.4 ± 1.3% of predicted value (p < 0.001), and the mean P axis decreased to 47.8 ± 1.6° (p < 0.001).

Sixty-two of 102 subjects (61%) had a vertical P axis; 26 (25%) had P axis in the intermediate range, and 14 (14%) had horizontal P axis (Figure 1).

View larger version (11K): [in this window] [in a new window]   Figure 1.   Distribution of mean P wave axis in all subjects before bronchodilators, and change in P axis after bronchodilators, n = 102.

Twenty-three of 27 subjects (85%) with TGV  130% predicted had a P axis  60°, as compared with 38 of 75 (51%) of those with TGV < 130% predicted (p < 0.005). A vertical P axis had a sensitivity of 85%, a specificity of 49%, a positive predictive value of 38% and a negative predictive value of 90% for the presence of significant hyperinflation.

Sixty-seven subjects (66%) had a decrease in the mean frontal P axis after bronchodilator, 21 (20%) showed no change, and 14 (14%) had an increase in the P axis (Figure 1). The mean decrease in P axis was 7.1 ± 1.1° as compared with a 1.12 ± 1.28° decrease in the QRS axis (p < 0.005). In subjects with a decrease in P axis, the mean decrease was 12.29 ± 2.21°.

Seventy-six subjects had  15% decrease in the TGV post-bronchodilator. Sixty two of 76 (82%) of these subjects had an associated significant decrease in P axis with a mean decrease of 10.49 ± 0.31 degrees. In comparison, only five of 26 (19%) of those who had less than 15% decrease in TGV had a decrease in P axis (p < 0.001). Sixty-two of 67 (93%) subjects with a decrease in P axis also had  15% decrease in TGV. A decrease in mean P axis predicted a decrease in lung volumes (sensitivity 82%, specificity 81%, positive predictive value 93%, negative predictive value 60%).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Electrocardiographic changes in acute and chronic lung disease in adults have been well recognized since 1935 when Winternitz first described the P pulmonale (1). Zuckerman and coworkers in 1948 demonstrated verticalization of the mean frontal P vector in chronic obstructive pulmonary disease (2). Since then, several investigators have studied ECG correlates with lung function in chronic lung disease (3). Spodick had shown that using P wave axis and configuration criteria, diffuse lung disease could be predicted by a blinded investigator (in 100 subjects) up to an accuracy of 93% (3). Spodick and coworkers (4) studied 301 adults older than 30 yr of age with pulmonary emphysema and compared ECGs with the 2-s fraction of the forced vital capacity. They demonstrated verticalization of the P axis with increased obstruction, and P-pulmonale correlated strongly with advanced obstruction.

Silver and colleagues (5) using regression analysis of rightward shift of P axis decrease in R amplitude in chest lead V6 and increase in T amplitude on ECG, could diagnose severe emphysema (FEV₁ < 50% of predicted) with 90% accuracy without any false positives. However their technique could not diagnose minimal or moderately advanced emphysema.

Shah and colleagues (8) studied 39 adults with chronic lung disease and compared the level of the right diaphragmatic leaf on chest radiography with the mean frontal P axis on ECG. 19 of 20 subjects with obstructive disease had a vertical P axis (70-90°), compared with 2 of 19 subjects with restrictive disease. All subjects with verticalization of the P axis had low diaphragm levels on chest radiographs. Previously Shah and coworkers (9) had shown that in a group of 30 adults with restrictive lung disease only four (13%) had a vertical P axis (70-90°) as compared with 77% of historical controls with obstructive disease.

All these studies involved adults with advanced degrees of obstruction. Similar to our findings, a vertical P axis correlated strongly with airway obstruction. There are no reported studies to date in the pediatric literature examining the relationship between P axis and lung function. We compared electrical axes on ECG with thoracic gas volumes derived by body plethysmography as a measure of hyperinflation in children with reversible airflow obstruction.

Verticalization of P axis was most marked in subjects with moderate to severe hyperinflation. P axis  60° had high sensitivity and excellent negative predictive value for the presence of significant hyperinflation. Furthermore a decrease in P axis of  5° was highly sensitive and predictive of a decrease in lung volumes.

If the vertical P axes were related to hyperinflation, we reasoned that a decrease in volumes with bronchodilator would alter the P axis. Indeed, 66% of the total subjects, and 69% of those with an initial vertical P axis had a subsequent decrease in P axis. Interestingly, there was no significant change in QRS axis, suggesting that atrial and not ventricular orientation is altered.

Our study group included children with mild airway obstruction on average. We speculate that the P wave axis would be even more predictive in a group of children presenting with severe airway obstruction, e.g., in an emergency room setting. We used 2.5 mg of albuterol, which is the standard FDA-approved dose, to assess for response to bronchodilators. A larger dose may have resulted in a greater response. We restricted our study to subjects  6 yr of age and able to perform spirometry reproducibly. We also measured both ECGs and lung volumes in subjects sitting upright. Previous studies in adults do not specify posture in which ECGs were obtained. We cannot extrapolate our findings to infants and young children. In this age range patients would be studied in the supine position. There is no data on effects of posture on the P axis in children.

In our study, as in previously reported studies, we calculated P axis manually. However most current computerized ECG machines are programmed to report cardiac axes including P axis, which would make this information easy for clinicians to obtain.

We speculate that studies in infants and young children may reveal similar findings. The P wave axis as a marker for increased lung volumes may be particularly useful in this age group since hyperinflation is so common even with mild degrees of obstruction, auscultation may be unremarkable and pulmonary function testing can be difficult. This technique could prove to be a practical, inexpensive and quick tool to assess hyperinflation and the response to bronchodilators.

We recommend a two-step method to predict hyperinflation in children using the ECG. If the initial P axis is < 60°, it is very unlikely that the patient has significant hyperinflation (negative predictive value of 90%) and a repeat ECG is not necessary. If the initial P axis is  60°, then a decrease in P axis of 5° after bronchodilator will confirm the presence of hyperinflation (positive predictive value of 93%).