Prevalence and Clinical Implications of Improper Filter Settings in Routine Electrocardiography Academic Article uri icon


MeSH Major

  • Diagnostic Errors
  • Electrocardiography
  • Heart Diseases


  • High- and low-filter bandwidth governs the fidelity of electrocardiographic waveforms, including the durations used in established criteria for infarction, the amplitudes used for the diagnosis of ventricular hypertrophy, and the accuracy of the magnitudes of ST-segment elevation and depression. Electrocardiographs allow users to reset high- and low-filter settings for special electrocardiographic applications, but these may be used inappropriately. To examine the prevalence of standard and nonstandard electrocardiographic filtering at 1 general medical community, 256 consecutive outpatient electrocardiograms (ECGs) submitted in advance of ambulatory or same-day admission surgery during a 3-week period were examined. ECGs were considered to meet standards for low-frequency cutoff when equal to 0.05 Hz and to meet standards for high-frequency cutoff when equal to 100 Hz, according to American Heart Association recommendations established in 1975. Only 25% of ECGs (65 of 256) conformed to recommended standards; 75% of ECGs (191 of 254) did not. The most prevalent deviation from standard was reduced high-frequency cutoff, which was present in 96% of tracings with nonstandard bandwidth (most commonly 40 Hz). Increased low-frequency cutoff was present in 62% of ECGs in which it was documented. In conclusion, improper electrocardiographic filtering, with potentially adverse clinical consequences, is highly prevalent at 1 large general medical community and is likely a generalized problem. This problem should be resolvable by targeted educational efforts to reinforce technical standards in electrocardiography.

publication date

  • March 2007



  • Academic Article



  • eng

Digital Object Identifier (DOI)

  • 10.1016/j.amjcard.2006.09.123

PubMed ID

  • 17317378

Additional Document Info

start page

  • 711

end page

  • 3


  • 99


  • 5