Objective-To estimate the cardiac output produced by external cardiac compression during standard cardiopulmonary resuscitation performed by two groups of operators with different levels of experience and training. Methods-Cardiac output was measured by Doppler aortovelography. All patients included in the study had necropsy examinations. Only patients without evidence of pulmonary embolism, myocardial rupture, aortic valve disease, or acute depletion of the intravascular volume were included. Results-31 patients presenting to the accident and emergency department suffering from non-traumatic cardiac arrest had cardiac output measurements made during resuscitation. Eleven patients were excluded after necropsy examination. The median cardiac index for the 20 study patients was 3.2 L min-' m-'. The Cardiac output can be estimated noninvasively using Doppler aortovelography.9 In this technique, a beam ofhigh frequency sound is reflected back from moving blood cells. The frequency of the reflected sound varies with the velocity and direction of the target.'0 From the difference in the frequency of the incident and reflected beams and the cross sectional area of the ascending aorta, the cardiac output can be estimated. In most cases the cross sectional area of the aorta cannot be accurately measured, so alternative Doppler indices of cardiac output have been used including systolic velocity integral or stroke distance. These Doppler indices have been shown to reflect cardiac output. " In this study, necropsy examinations after unsuccessful resuscitation allowed measurement of aortic diameter so that cardiac output could be assessed more accurately.The aim of our study was to compare the cardiac output produced by external cardiac compression when this was performed by personnel with different levels of experience and training. We also compared the degree of resuscitation trauma associated with different operators. Methods PATIENTSPatients who had suffered a non-traumatic cardiorespiratory arrest were eligible for the study. Standard cardiopulmonary resuscitation was carried out in accordance with the Resuscitation Council guidelines.'2 The patients were intubated and ventilated with 100% oxygen by hand using a Water's circuit. External cardiac compression was performed at a rate of 80/min and the lungs inflated after every fifth compression. The massage was performed by one of the three personnel, each of whom had exceeded the criteria for a pass result of the Resuscitation Examination for Membership of the Royal College of General Practitioners." In the case of a cardiorespiratory arrest, it is accepted practice to rotate personnel to perform external cardiac massage and this minimises operator fatigue. In order to assess whether the cardiac output was operator dependent, additional measurements were made on seven cases, with the massage being performed by less experienced operators. These personnel (student nurses) had been trained in basic life support and were familiar with the technique, having used ma...