Agreement andreproducibility oftheestimates ofcardiovascular function byimpedance cardiography andM-mode echocardiography inhealthy subjects

Thereproducibility andagreementoftheestimates ofstroke volume (SV), cardiac output (CO)andtotal peripheral resistance (TPR)bytransthoracic impedance cardiography (ZCG)andM-modeechocardiography (ECHO)were analyzed before andafter the placebo-controlled administration ofascending doses ofisosorbide dinitrate andnicorandil in12healthy subjects. There was no biostatistical agreement between thetwomethods inestimating cardiovascular function either before or after dosing (ZCGestimated substantially larger SV,CO andlower TPR).But,ZCG andECHO estimated about similar overall treatment related changes (across treatments andperiods) andreached substantially better agreementwhenthevalues were expressed asratio ofthebaseline before dosing. Suchimprovement didnotoccurwhenthedatawere expressed as arithmetic difference frombaseline. Inspite oftheimprovement ofagreementby expressing thedata asratio ofbaseline, thecoefficients ofreproducibility between the two methods(circa 25% ofbaseline) remained toolarge tojudgethemethods interchangeable.