AIM: To investigate the utility of transthoracic echocardiography in confirming appropriate pulmonary artery catheter(PAC) placement. METHODS: Three commonly used transthoracic echocardiography(TTE) views were used to...AIM: To investigate the utility of transthoracic echocardiography in confirming appropriate pulmonary artery catheter(PAC) placement. METHODS: Three commonly used transthoracic echocardiography(TTE) views were used to confirm PAC position in 103 patients undergoing elective cardiac surgery- the parasternal short axis right ventricular inflow-outflow view; the subcostal short axis right ventricular inflow-outflow view; and the parasternal short axis ascending aortic view. All PACs were inserted by the managing anesthesiologist under pressure waveform guidance alone, who was blinded to all sonographic information. A sonographer blinded to all pressure waveform information confirmed visualisation of an "empty" PA before PAC insertion, and visualisation of the PAC balloon in the main PA(MPA) or right PA(RPA) after attempts at placement were complete. Agreement, sensitivity and specificity of TTE in confirming appropriate PAC placement was compared against pressure waveformguidance as the "gold standard". The successful view used was compared against patients' anthropomorphic indices, presence of lung hyperinflation, and insertion of PAC during positive pressure ventilation. Agreement between TTE and pressure waveform guidance was analysed using Cohen's Kappa statistic. The relative proportion of total RPA seen by subcostal vs parasternal TTE views was also compared with a further 20 patients' computed tomography(CT) pulmonary angiograms(CTPA), to determine efficacy in detection of distal RPA PAC placement. RESULTS: Appropriate positioning of the PAC balloon, and its to-and-fro movement consistent with a nonwedged state, within the MPA or RPA was confirmed by TTE in 98 of the 103 patients [sensitivity 95%(95%CI: 89%-98%)], and absence of the PAC balloon before insertion correctly established in 100 patients [specificity 97%(92%-99%)]. This was in very good agreement with pressure waveform guidance [Cohen's Kappa 0.92,(0.87-0.98)]. The subcostal view was the best view to visualise the PAC tip when it was placed in the right pulmonary artery(OR 70, P < 0.0001), was more successful in patients with COAD(OR 9.5, P = 0.001), and visualized 61%(vs 44% by parasternal views, P < 0.001) of mean RPA lengths compared with CTPA; however the parasternal views were more successful in patients with higher body mass indexs(OR 0.78 for success with subcostal views, P < 0.001). There was a trend towards insertion during intermittent positive pressure ventilation favoring visualisation by subcostal views(OR 3.9, P = 0.08). The subcostal view visualized a greater length of the RPA than parasternal views(3.9 cm vs 2.9 cm, P < 0.0001). PACs were more often placed in the MPA than RPA(80 vs 18 patients). Three patient's pulmonary arteries were not visible by any TTE view; in a further 2 patients, despite preinsertion visualisation of their pulmonary arteries, the PAC balloon was not visible by any view with TTE where correct placement by pressure waveform was unequivocal.展开更多
文摘AIM: To investigate the utility of transthoracic echocardiography in confirming appropriate pulmonary artery catheter(PAC) placement. METHODS: Three commonly used transthoracic echocardiography(TTE) views were used to confirm PAC position in 103 patients undergoing elective cardiac surgery- the parasternal short axis right ventricular inflow-outflow view; the subcostal short axis right ventricular inflow-outflow view; and the parasternal short axis ascending aortic view. All PACs were inserted by the managing anesthesiologist under pressure waveform guidance alone, who was blinded to all sonographic information. A sonographer blinded to all pressure waveform information confirmed visualisation of an "empty" PA before PAC insertion, and visualisation of the PAC balloon in the main PA(MPA) or right PA(RPA) after attempts at placement were complete. Agreement, sensitivity and specificity of TTE in confirming appropriate PAC placement was compared against pressure waveformguidance as the "gold standard". The successful view used was compared against patients' anthropomorphic indices, presence of lung hyperinflation, and insertion of PAC during positive pressure ventilation. Agreement between TTE and pressure waveform guidance was analysed using Cohen's Kappa statistic. The relative proportion of total RPA seen by subcostal vs parasternal TTE views was also compared with a further 20 patients' computed tomography(CT) pulmonary angiograms(CTPA), to determine efficacy in detection of distal RPA PAC placement. RESULTS: Appropriate positioning of the PAC balloon, and its to-and-fro movement consistent with a nonwedged state, within the MPA or RPA was confirmed by TTE in 98 of the 103 patients [sensitivity 95%(95%CI: 89%-98%)], and absence of the PAC balloon before insertion correctly established in 100 patients [specificity 97%(92%-99%)]. This was in very good agreement with pressure waveform guidance [Cohen's Kappa 0.92,(0.87-0.98)]. The subcostal view was the best view to visualise the PAC tip when it was placed in the right pulmonary artery(OR 70, P < 0.0001), was more successful in patients with COAD(OR 9.5, P = 0.001), and visualized 61%(vs 44% by parasternal views, P < 0.001) of mean RPA lengths compared with CTPA; however the parasternal views were more successful in patients with higher body mass indexs(OR 0.78 for success with subcostal views, P < 0.001). There was a trend towards insertion during intermittent positive pressure ventilation favoring visualisation by subcostal views(OR 3.9, P = 0.08). The subcostal view visualized a greater length of the RPA than parasternal views(3.9 cm vs 2.9 cm, P < 0.0001). PACs were more often placed in the MPA than RPA(80 vs 18 patients). Three patient's pulmonary arteries were not visible by any TTE view; in a further 2 patients, despite preinsertion visualisation of their pulmonary arteries, the PAC balloon was not visible by any view with TTE where correct placement by pressure waveform was unequivocal.
