To examine the relationship of pulmonary artery catheter
(PAC) use to patient outcomes, including mortality
rate and resource utilization
, in patients with severe sepsis
in eight academic medical centers.
Case-control, nested within a prospective cohort study
Eight academic tertiary care centers.
Stratified random sample of 1,010 adult admissions with severe sepsis
Measurements and Main Results
The main outcome measures were in-hospital mortality
, total hospital charge, and length of stay (LOS) for patients with and without PAC use. The case-matched subset of patients included 141 pairs managed with and without the use of a PAC. The mortality
rate was slightly but not statistically significantly lower among the PAC use group compared with those not using a PAC (41.1% vs. 46.8%, p
= .34). Even this trend disappeared after we adjusted for the Charlson comorbidity score and sepsis
-specific Acute Physiology and Chronic Health Evaluation (APACHE) III (adjusted odds ratio, 1.02; 95% confidence interval, 0.61–1.72). In linear regression models adjusted for the Charlson comorbidity score, sepsis
-specific APACHE III, surgical status, receipt of a steroid before sepsis
onset, presence of a Hickman catheter
, and preonset LOS, no significant differences were found for total hospital charges (US$139,207 vs. 148,190, adjusted mean comparing PAC and non-PAC group, p
= .57), postonset LOS (23.4 vs. 26.9 days, adjusted mean, p
= .32), or total LOS in intensive care unit (18.2 vs. 18.8 days, adjusted mean, p
Among patients with severe sepsis
, PAC placement was not associated with a change in mortality
rate or resource utilization
, although small nonsignificant trends toward lower resource utilization
were present in the PAC group.