Pulmonary artery catheterization: a narrative and systematic critique of randomized controlled trials and recommendations for the future.

Objective: The purpose of this review is to evaluate the current literature regarding the efficacy of pulmonary artery catheterization, to assess the quality of existing randomized controlled trials and to make recommendations for future studies.

Data source: A Medline search was conducted of English language studies published between 1970 and 1996 using as search terms "Swan-Ganz catheterization," "pulmonary artery catheterization," and "right heart catheterization."

Study selection: Randomized controlled trials of pulmonary artery catheterization were selected for review. Sixteen randomized controlled trials were identified.

Data extraction: The quality of the randomized controlled trials was assessed and a formal meta-analysis was performed.

Data synthesis: The overall quality score was 40.15 +/- 6.32 out of a possible score of 100. Serious deficiencies were identified including a lack of a priori sample size calculations, unclear definitions of concomitant therapy, inability to blind physicians and patients, and lack of blinded outcome assessments. A random effects model found the relative risk ratio of .808 (95% confidence interval [CI] = .598, 1.091; p = .148). An analysis of potential covariates revealed patients from surgical series treated with pulmonary artery catheter (PAC)-guided therapy had a relative risk ratio of .578 (95% CI = .357, .937; p = .03) compared with control groups. Patients from medical or mixed unit series had a relative risk of 1.043 (95% CI = .784, 1.387; p = NS).

Conclusion: Existing randomized controlled trials on PAC-guided strategies reveal a modest risk reduction that does reach statistical significance. Risk reduction appears to be greatest in surgical series. Deficiencies of these trials have important implications for the proper design of future trials.