REDUCING ERRORS IN SEGMENTAL ANALYSIS (SA) FOR PEDIATRIC ECG USING A COMPLEMENTARY SOFTWARE

Molina M.S., Benjó A.M., Favarato D., Tobias N.M., Nigri M., Chagas A.C.P.,
Ramires J.A.F., Pastore C.A.
Heart Institute (InCor) of the University of São Paulo Medical School
Av. Dr. Enéas de Carvalho Aguiar, 44 – 05403-000 São Paulo, SP – Brazil

INTRODUCTION

Several components can cause imprecision of ECG segments classification in pediatric electrocardiography and affect accuracy of the medical report. The conventional memory analysis or quick reference to a standard table of reference may lead to mistakes, therefore to a biased final medical report.

OBJECTIVE

To evaluate the accuracy of the use of computer software in aiding electrocardiographic diagnosis in infant patients.

MATERIAL AND METHODS

Five standard ECGs were independently analyzed twice by nineteen cardiology experts, who fulfilled 1,140 segmental analyses (SA) per group. At first, ECGs were conventionally examined using the normality reference table. Subsequently the data, generated by a manual and systematic measurement of ECGs, were fed into the computer software. SA is defined as the measurement or forecast of twelve specific segments.

RESULTS

Table I shows elements of analysis and results.

Performing the analysis using the software resulted in an 83% reduction in the chance of reading errors on interpreting the SA; the potentially damaged analysis of ECGS was reduced by 78%; and the potentially damaged analysis of ECGS when the software plus the conventional method were used was reduced by 94%.

CONCLUSION

This support software has provided evidence of its efficacy and safety as a complementary tool for electrocardiographic diagnosis in Pediatric Cardiology.

TABLE I

Elements of analysis	Control	Software	p
Mistakes in segmental analysis	95(  8.3%)	16(  1.4%)	0.003
ECGs with potentially damaged analysis	36( 37.9%)	8(  8.0%)	0.001
... and after complementary method	2(-94.4%)	1(-87.5%)	0.001