Facebook Twitter YouTube


Use of a Standardized Approach to Evaluation of Pediatric Chest Pain: Does It Make a Difference?

Friday, October 19, 2012
Room 275-277 (Morial Convention Center)
BM Cardis, PA Frias, HM Phelps, DM Drossner and ME Oster, Sibley Heart Center Cardiology/Children's Healthcare of Atlanta/Emory University, Atlanta, GA

Purpose: Chest pain remains a common reason for referral to pediatric cardiologists.   Within our group, thirty-six physicians evaluate approximately two thousand children annually with the chief complaint of chest pain. In October 2010, our practice instituted a standardized evaluation protocol in an attempt to develop a comprehensive yet cost-effective method to identify cardiac causes of chest pain in children and to limit practice variability amongst physicians.

Methods: Beginning in October 2010, all patients evaluated in an outpatient clinic with a chief complaint of chest pain were required to complete a standardized “Chest Pain Questionnaire” which included objective answers to descriptive questions regarding chest pain.   In addition, patients completed a “Cardiac Risk Assessment Form” in order to identify risk factors for sudden cardiac death. This data was used in conjunction with ECG (performed on all patients) and physical examination.  Abnormalities on the chest pain questionnaire, cardiac risk assessment form, ECG, or examination warranted further evaluation with echocardiogram according to the protocol.    We retrospectively reviewed data from all new patients seen for a chief complaint of chest pain from March 2010 to December 2011. We used chi-square analysis to compare echocardiogram utilization rates prior to initiation of the protocol (period 1: March – September 2010) to rates after the initiation of the protocol (period 2: October 2010 – December 2011).

Results:  There were 2416 patients seen with a chief complaint of chest pain during the protocol period (October 2010 – December 2011) with age range 2-26 years (median=13 years).    Abnormal findings on either ECG or echocardiogram occurred in 97 patients (4%).  A cardiac etiology for the chest pain was found in 21 patients (0.9%).  Potentially life-threatening cardiac conditions were identified in 13 patients (0.5%), including anomalous left coronary artery (n=1), hypertrophic cardiomyopathy (n=1), Wolff-Parkinson-White syndrome (n=5), mitral valve prolapse (n=5), and a combined finding of both Wolff-Parkinson-White syndrome and mitral valve prolapse (n=1).   Echocardiogram was utilized in 1439 /2416 patients (59.6%) compared with 561/1053 (53.6%) prior to protocol initiation (P<0.001).  Despite an overall increase in echocardiogram usage, this finding was not consistent throughout the group: 9/22 physicians who evaluated more than 20 patients with chest pain during the initial period had a lower utilization of echocardiogram after initiation of the protocol. 

Conclusion:  Cardiac etiology of chest pain remains rare.  However, life-threatening cardiac conditions do present as chest pain in children. Identification of these patients is imperative.  Development of a cost-effective and efficient method of identifying this small subset of patients is difficult but not impossible. Further detailed analysis of our chest pain questionnaire and cardiac risk assessment form may result in the ability to assign predictive values to each response with an eventual decrease in echocardiogram utilization.