Effects of electrocardiograph frequency filters on P-QRS-T amplitudes of the feline electrocardiogram.

OBJECTIVE--To determine whether standard manual frequency filters in the ON and OFF settings affected P-QRS-T voltages, discover whether recorded P-QRS-T voltages vary between commercial electrocardiographs, assess effects of frequency filters on base-line artifact, and evaluate ECG frequency content by high-fidelity recordings subjected to digital filters with variable frequencies. DESIGN--Sequential 10-lead ECG were recorded in 30 cats, using 3 commercial electrocardiographs to assess effects of manual frequency filters on the P-QRS-T wave forms. Three clinically normal cats were evaluated for ECG frequency content. ANIMALS--Thirty cats (13 with hypertrophic cardiomyopathy; 4 with restrictive cardiomyopathy; 3 hyperthyroid; 1 with ventricular septal defect; 1 with aortic stenosis; and 8 with no detectable cardiovascular disease). Three additional clinically normal cats were studied for effects of frequency filters on the ECG frequency content. PROCEDURES--Ten-lead ECG were recorded on each cat by use of 3 commercial electrocardiographs sequentially. For each machine, a recording was made with manual filters ON, immediately followed by a recording with manual filters OFF. High-fidelity lead-II ECG recordings were made with filters set with their rolloff frequency at 0.1 Hz and 3.0 kHz; output voltage (0.2 mV/V) was fed to an analog-to-digital converter, then to attendant software, which sampled the signal at 6 kHz with a 12-bit sampler, and were digitally filtered at various corner frequencies. RESULTS--Voltages recorded by all 3 electrocardiographs were greatest when filters were OFF (most prominent on R- and S-wave voltages). In all recorded leads, R-wave voltage was significantly greater when filters were OFF than ON. Comparison of voltages indicated significant (P < 0.05) differences between R-wave voltages recorded in all leads with manual filters ON, but not with filters OFF. With filters ON, each electrocardiograph produced a smaller percentage of recordings with moderate to severe baseline artifact than with filters OFF. R-Wave amplitudes of high-fidelity lead-II ECG were significantly decreased with digital filters set at corner frequencies < 150 Hz. CONCLUSIONS--Significant (P < 0.05) voltage attenuation was recorded by each of the 3 commercial electrocardiographs when frequency filters were ON, compared with OFF. Comparison of waveform voltages among electrocardiographs with filters ON indicated significant variation in R-wave amplitudes in all leads. With manual filters ON, each electrocardiograph recorded a smaller percentage of recordings with baseline artifact than with filters OFF. Substantial frequency components > or = 150 Hz are present in the feline ECG waveform. Thus, filters with frequencies < 150 Hz markedly attenuate the feline R wave. CLINICAL RELEVANCE--Attenuation of feline ECG signals occurs with use of commercial electrocardiographs and varies greatly between manufacturers. This is attributable largely to internal manual frequency filters. These consequences may be important when applying standard feline reference values or when equivocal voltage measurements are recorded.