Recommendations for ECG diagnostic coding

Abstract

The Oxford dictionary defines code as "a body of laws so related to each other as to avoid inconsistency and overlapping". It is obvious that natural language with its high degree of ambiguity does not qualify as a code in the sense of this definition. Everyday experiences provide ample evidence that natural language, because of its richness and lack of uniqueness, is subject to multiple interpretations and thus not suitable for conveying ideas or data in an unequivocal, uniform and concise manner. For this reason codes have been developed and used in several areas of medicine [1-3] to describe, document, and transmit qualitative medical data. It is rather surprising that electrocardiography has been able to exist for so long without any formalized language to describe its findings. Increased use of electrocardiograms in epidemiology, large scale electrocardiographic studies and last but not least computerized EeG interpretation have provided incentives to develop codes. Initial efforts in this direction [4-6] were primarily guided by loc al needs for improved storage, retrieval and handling of information; without major modifications they do not, however, satisfy all the requirements one expects from an EeG code today. Nevertheless, the experience gained in the use of these early EeG codes provides an important source of information on which to build specifications for a new or expanded code. It is significant that several members of this working group have been extensive users of the Booth-Hull code [4] and the Utrecht coding system [5] .