Review: a pulmonary artery catheter does not reduce mortality or hospital days in critically ill patients.

M e t h o d s Data sources: MEDLINE (1985 to 2005), the Cochrane Central Register of Controlled Trials (1988 to 2005), the National Institutes of Health Clinical Trials Web site, the U.S. Food and Drug Administration Web site, and references of retrieved articles. Study selection and assessment: Randomized controlled trials (RCTs) that compared PAC with no PAC and assessed death, days in hospital, or days in the intensive care unit. Studies that combined randomized and nonrandomized groups in outcomes, did not specify groups as PAC or no PAC, or did not use conventional PAC were excluded. Study quality was assessed for allocation generation, allocation concealment, and intention-totreat analysis. Outcomes: All-cause mortality during hospitalization and number of hospital days.