The use of a pulmonary artery catheter (PAC) -- a device used for more than 30 years to assess cardiovascular health and to help guide the treatment of critically ill patients -- does not appear to improve outcomes nor confer added risks to patients. The finding came from a new meta-analysis conducted by the Duke Clinical Research Institute (DCRI) of data collected from 13 randomized clinical trials studying the device.

The researchers, who reported the results of their analysis in the Oct. 5, 2005, issue of the Journal of the American Medical Association (JAMA), concluded that PACs should not be used routinely in patients in intensive care units (ICU), patients with severe heart failure or patients undergoing surgery, until further studies can uncover any correlations between outcomes from the use of the device with specific treatments.

In using the device, physicians insert a catheter into the venous system around the heart to obtain information about a variety of hemodynamic parameters, such as filling pressures, pumping strength of the heart, saturation levels of oxygen in the blood and how much fluid is being retained. The procedure is usually performed at the bedside, and the catheter can be left in place for a number of days to provide constant monitoring. More than 1 million PAC procedures are performed each year in the U.S.

"The main question this study raises is where the use of PAC fits into the overall therapeutic plan for these patients," said cardiologist Monica Shah, M.D., first author of the paper. She led the team while at the DCRI before leaving Duke for Columbia University Medical Center. "The PAC is a diagnostic tool, like an x-ray or an electrocardiogram, and as such, we don't necessarily expect them to improve outcomes on their own.

"Our analysis showed that PAC doesn't improve outcome as a part of a therapeutic strategy," she continued. "Maybe the catheter could be helpful if, for example, it guided us in selecting the right medications. To date, those types of trials have not been conducted."

The researchers conducted the meta-analysis as a way of putting into perspective the results of another trial - dubbed ESCAPE. That results of that trial, also being published in the Oct. 5, 2005, issue of JAMA, found that while the use of PAC is just as safe as non-invasive methods of assessing the cardiovascular status of patients with heart failure, its use did not improve clinical outcomes. That trial was led by the DCRI and researchers at Brigham & Women's Hospital, Boston.

For the current meta-analysis, the team collected patient data from 13 randomized clinical trials, including ESCAPE. A total of 5,051 patients were involved. The trials enrolled patients with advanced heart failure, those awaiting major surgery or those who were admitted to ICUs for a spectrum of serious illnesses.

Despite more than 30 years of physician experience with PAC, there is still much debate about the device's utility and where it belongs in the care of patients. According to the researchers, there have been many small trials of PAC use in different medical settings that have not provided clear guidance on the device's efficacy or safety.

"To help put the results of ESCAPE into context, we wanted to look at how the PAC performed in different settings and situations in a wide range of patients," said the DCRI's Vict Hasselblad, Ph.D., lead statistician for the meta-analysis. Hasselbald explained that the patients enrolled in the ESCAPE trial had to be at what physicians call medical equipoise, meaning that in the physician's mind, each patient had an equal chance of either benefiting or not benefiting from the procedure. For example, patients were excluded from the trials if their physicians felt that they were so extremely ill that PAC was absolutely necessary, or if the patients were considered too healthy.

"For that reason, we concluded that the routine use of the PAC should be avoided for patients in the ICU or with heart failure," Hasselblad said. "Its use should be saved for those patients who absolutely need it."

Shah pointed out that there may be an important role for PAC to play, especially in those patients who were not included in the previous randomized clinical trials, especially those who are considered candidates for heart or lung transplantation.

"What is definitely needed is a clinical trial that looks at PAC in conjunction with specific drugs that are associated with its use," Shah said. "That way we could really see if there is any impact."

The researchers did find in their meta-analysis that those patients who were randomized to PAC had significantly higher usage rates of drugs that dilate blood vessels, as well as drugs that stimulate the contraction of heart muscle. However, they also found that in those patients, there was no significantly increase in mortality, survival or days spent in the hospital