[Diagnostic and therapeutic consequences of pulmonary artery catheters. Prospective evaluation in an intensive care unit].

BACKGROUND

The diagnostic value and the therapeutic impact of pulmonary artery catheters have been repeatedly challenged. As part of a quality assurance project we compared clinical assessment with invasive measurements of hemodynamic parameters in critically ill patients and assessed changes in therapy following catheter insertion.

METHODS

In 47 consecutive patients (age 31-79 years; mean APACHE-II score 19 +/- 8) of a medical intensive care unit both physicians and intensive care nurses independently estimated mean pulmonary arterial pressure (PAPm), pulmonary capillary wedge pressure (PCWP), and cardiac output (CO) prior to pulmonary artery catheterization based on available clinical and radiological information. The physician-in-charge had to specify a presumptive treatment strategy in case hemodynamic measurements were not available.

RESULTS

In 19% minor complications due to catheter insertion occurred, only one requiring intervention (intravenous adenosine). Overall, PAPm was correctly predicted in 56%, PCWP in 51% and CO in 50% of the patients (Z > 2.6; p < 0.01 for all 3 parameters). There was no statistically significant difference in the percentage of correct predictions between physicians of different training levels or between physicians and nurses. In patients with sepsis the clinical prediction of PAPm (37% correct) and of PCWP (37%) was probably not better than by chance alone (Z = 1.53; p = 0.06). In 21% there was a major change in treatment after obtaining the invasive hemodynamic values compared to the preinsertion plan.

CONCLUSIONS

Although randomized trials have not yet shown pulmonary artery catheterization to decrease mortality in critically ill patients, possible useful effects on intermediate outcomes cannot be excluded in view of the high numbers of misjudgements based on clinical information alone and the considerable number of changes in therapy following catheter insertion. Reducing uncertainty in clinical judgement may not in every case lead to a beneficial strategy, but it may prevent potentially harmful decisions.