Increasing use of less-invasive hemodynamic monitoring in 3 specialty surgical intensive care units: a 5-year experience at a tertiary medical center.

INTRODUCTION

Less-invasive hemodynamic monitoring (eg, esophageal doppler monitoring [EDM] and arterial pressure contour analysis, FloTrac) is increasingly used as an alternative to pulmonary artery catheters (PACs) in critically ill intensive care unit (ICU).

HYPOTHESIS

The decrease in use of PACs is not associated with increased mortality.

METHODS

Five-year retrospective review of 1894 hemodynamically monitored patients admitted to 3 surgical ICUs in a university-affiliate, tertiary care urban hospital. Data included the number of admissions, diagnosis-related group discharge case mix, length of stay, insertion of monitoring devices (PAC, EDM, and FloTrac probes), administered intravenous vasoactive agents (β-predominant agonists--dobutamine, epinephrine, and dopamine; vasopressors--norepinephrine and phenylephrine), and mortality. Data from hospital administrative databases were compiled to create patient characteristic and monitoring variables across a 5-year time period, 2005 to 2009 inclusive. Chi-square for independent proportions, 1-way analysis of variance, and Kruskal-Wallis tests were used; tests for trend were conducted. An α level of .05 was considered significant. Statistical Package for the Social Sciences v14 was used for all statistical testing.

RESULTS

There was a significant change in the type of hemodynamic monitors inserted in 2 of the 3 surgical ICUs (in the general surgery and neurointensive care but not in the cardiac ICU) from PACs to less-invasive devices (FloTrac or EDM) during the 5-year study period (P < .001). There was no change in mortality rate over the time period (P = .492). There was an overall increase in the proportion of monitored patients who received intravenous vasoactive agents (P < .001) with a progressive shift from β-agonists to vasopressors (P < .002). Multivariate analyses indicated that age, case mix, and use of vasoactive agents were all independent predictors of inhospital mortality (P = .001) but that type of monitoring was not (P = .638).

CONCLUSIONS

In a 5-year period, the decreased insertions of PACs were replaced by increased utilization of less-invasive hemodynamic monitoring devices. This change in practice did not adversely impact mortality.