BACKGROUND AND OBJECTIVES: To analyse the precision of transpulmonary thermodilution from the PiCCO technique (Pulsion Medical System, Munich, Germany) in everyday intensive care practice in order to ascertain the minimum number of measurements necessary for scientific precision. METHODS: An observational study in the medical-surgical ICU of a teaching hospital was performed. Thirty consecutive patients from a mixed intensive care population using the PiCCO haemodynamic monitor were included. Five thermodilution measurements were repeated at 2 min intervals. The variability of the cardiac index and the global end-diastolic volume index was analysed with respect to the five consecutive measurements and the mean of the first two, first three, first four and all five measurements. RESULTS: There was similar distribution among the different measurements and means. The variability of the cardiac index and global end-diastolic volume index, represented by the standard error of means, the coefficient of errors and the confidence intervals, revealed a similar precision in separate measurements and in the different averaging techniques. The coefficient of errors was less than 5% even when calculating the mean of the first two measurements, meeting the criterion of scientific precision, and including patients with arrhythmia and varying blood pressure. CONCLUSION: Calculating the mean of two good-quality transpulmonary thermodilution measurements is equivalent to the other averaging techniques (three to five measurements) for the cardiac index and global end-diastolic volume index. Any further repeated measurements may be unnecessary and may contribute to volume overloading.
BACKGROUND AND OBJECTIVES: To analyse the precision of transpulmonary thermodilution from the PiCCO technique (Pulsion Medical System, Munich, Germany) in everyday intensive care practice in order to ascertain the minimum number of measurements necessary for scientific precision. METHODS: An observational study in the medical-surgical ICU of a teaching hospital was performed. Thirty consecutive patients from a mixed intensive care population using the PiCCO haemodynamic monitor were included. Five thermodilution measurements were repeated at 2 min intervals. The variability of the cardiac index and the global end-diastolic volume index was analysed with respect to the five consecutive measurements and the mean of the first two, first three, first four and all five measurements. RESULTS: There was similar distribution among the different measurements and means. The variability of the cardiac index and global end-diastolic volume index, represented by the standard error of means, the coefficient of errors and the confidence intervals, revealed a similar precision in separate measurements and in the different averaging techniques. The coefficient of errors was less than 5% even when calculating the mean of the first two measurements, meeting the criterion of scientific precision, and including patients with arrhythmia and varying blood pressure. CONCLUSION: Calculating the mean of two good-quality transpulmonary thermodilution measurements is equivalent to the other averaging techniques (three to five measurements) for the cardiac index and global end-diastolic volume index. Any further repeated measurements may be unnecessary and may contribute to volume overloading.